Highly efficient ligands for dihydrofolate reductase from Cryptosporidium hominis and Toxoplasma gondii inspired by structural analysis

Article abstract of DOI:10.1021/jm061027h

The search for effective therapeutics for cryptosporidiosis and toxoplasmosis has led to the discovery of novel inhibitors of dihydrofolate reductase (DHFR) that possess high ligand efficiency: compounds with high potency and low molecular weight. Detaile

Full text of DOI:10.1021/jm061027h

940
J. Med. Chem. 2007, 50, 940-950
Highly Efficient Ligands for Dihydrofolate Reductase from Cryptosporidium hominis and
Toxoplasma gondii Inspired by Structural Analysis
Phillip M. Pelphrey,†,‡ Veljko M. Popov,†,‡ Tammy M. Joska,§ Jennifer M. Beierlein,# Erin S. D. Bolstad,# Yale A. Fillingham,‡
Dennis L. Wright,|,#,* and Amy C. Anderson#,*
Department of Chemistry, Dartmouth College, HanoVer, New Hampshire 03755, Department of Biochemistry, Dartmouth Medical School,
HanoVer, New Hampshire 03755, Department of Chemistry, UniVersity of Connecticut, Storrs, Connecticut 06269, and Department of
Pharmaceutical Sciences, UniVersity of Connecticut, Storrs, Connecticut 06269
ReceiVed August 25, 2006
The search for effective therapeutics for cryptosporidiosis and toxoplasmosis has led to the discovery of
novel inhibitors of dihydrofolate reductase (DHFR) that possess high ligand efficiency: compounds with
high potency and low molecular weight. Detailed analysis of the crystal structure of dihydrofolate reductase-
thymidylate synthase from Cryptosporidium hominis and a homology model of DHFR from Toxoplasma
gondii inspired the synthesis of a new series of compounds with a propargyl-based linker between a substituted
2,4-diaminopyrimidine and a trimethoxyphenyl ring. An enantiomerically pure compound in this series exhibits
IC50 values of 38 and 1 nM against C. hominis and T. gondii DHFR, respectively. Improvements of 368-
fold or 5714-fold (C. hominis and T. gondii) relative to trimethoprim were generated by synthesizing just
14 new analogues and by adding only a total of 52 Da to the mass of the parent compound, creating an
efficient ligand as an excellent candidate for further study.
Introduction
Cryptosporidium and Toxoplasma are apicomplexan parasitic
protozoa that cause severe disease in the population worldwide.
Cryptosporidiosis, caused by C. hominis, is characterized by
wasting disease and most often affects immune-compromised
patients, the elderly, and day-care children, although very large
outbreaks have occurred in otherwise healthy populations.1
There is no effective therapy for cryptosporidiosis. Toxoplas-
mosis, when transmitted congenitally as T. gondii, can cause
neonatal death and when transmitted through ingestion of
contaminated meat or water, can cause fever and sore throat,
or cerebral inflammation in immune-compromised patients.1
Both of these parasitic protozoa have been classified as Category
Figure 1.
for the pathogenic form of the enzyme possible. The DHFR
inhibitor pyrimethamine (1, Figure 1) has been effectively used
to treat toxoplasmosis2 as well as malaria,3 caused by another
apicomplexan parasite, Plasmodium. However, many patients
have had severe reactions to pyrimethamine, limiting its efficacy.
Trimethoprim (2, TMP, Figure 1) has been used effectively
in the clinic as an antibacterial agent since the 1960s. It possesses
excellent drug-like characteristics including relatively low
molecular weight (MW ) 290). However, it exhibits high
affinity for only a small subset of species of DHFR from
pathogenic organisms such as Escherichia coli, thus limiting
its widespread application. TMP exhibits only moderate in Vitro
potency against DHFR from C. hominis (ChDHFR) (IC50 ) 14
µM)4 and T. gondii (TgDHFR) (IC50 ) 8 µM).5,6 Although TMP
may not be clinically useful against these DHFR targets, it may
serve as a lead compound in the design of higher potency DHFR
inhibitors.
It is often appreciated that during the lead optimization
process, increases in potency correlate with increases in mo-
lecular weight. However, the additional molecular weight
frequently represents a liability as it compromises the drug-
like properties of the lead.7,8 The necessary compromise between
increasing molecular weight and increasing affinity can be
examined quantitatively with the concept of ligand efficiency.9,10
Ligand efficiency is defined as the overall binding energy per
B biodefense agents.
Dihydrofolate reductase (DHFRa) has been a validated drug
target for treatment of protozoal infections for decades. Occur-
ring as a bifunctional protein with thymidylate synthase
(DHFR-TS) in the apicomplexan protozoa, DHFR utilizes the
cofactor NADPH to catalyze the reduction of dihydrofolate to
tetrahydrofolate, thereby performing a key reaction in the sole
de noVo synthesis of deoxythymidine monophosphate (dTMP).
The overall fold of DHFR is widely conserved throughout
evolution; however, there are several residue differences in the
active sites of different species that make achieving selectivity
* To whom correspondence should be addressed. Mailing address: Dept.
of Pharmaceutical Sciences, University of Connecticut, 69 N. Eagleville
Rd., Storrs, CT 06269. Phone (D.L.W.): (860) 486-1556, (A.C.A.): (860)
486-6145. Fax: (860) 486-6857. E-mail: Dennis.Wright@uconn.edu,
Amy.Anderson@uconn.edu.
‡ Department of Chemistry, Dartmouth College.
§ Department of Biochemistry, Dartmouth Medical School.
| Department of Chemistry, University of Connecticut.
# Department of Pharmaceutical Sciences, University of Connecticut.
† These authors contributed equally.
a Abbreviations: DHFR, dihydrofolate reductase; DHFR-TS, dihydro-
folate reductase-thymidylate synthase; dTMP, deoxythymidine monophos-
phate; ChDHFR, dihydrofolate reductase from Cryptosporidium hominis;
TgDHFR, dihydrofolate reductase from Toxoplasma gondii; TMP, tri-
methoprim; MTX, methotrexate; pABA, para-aminobenzoic acid; NADPH,
nicotinamide adenine dinucleotide phosphate.
10.1021/jm061027h CCC: $37.00 © 2007 American Chemical Society
Published on Web 02/02/2007

Highly Efficient Ligands for DHFR
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5 941
Figure 2. (a) Sequence alignment of ChDHFR (PDB code: AAB00163) and TgDHFR (PDB code: XP665866). Arrows indicate residues located
in the active site. (b) MTX modeled in the ChDHFR active site.
Scheme 1a
a Reagents and conditions: (a) ICl, MeOH, 25 °C, 15 h, then 1N NaOH, 25 °C, 2 h (80%); (b) Pd(PPh3)2Cl2, CuI, Et3N, DMF, 100 °C, 2 h (90%); (c)
5% Pd/C, EtOH, 45 psi H2, 25 °C, 5 h (33%, 81% borsm); (d) I2, THF, 25 °C, 30 min (80%); (e) 10% Pd/C, MeOH, 50 psi H2, 25 °C, 10 h (87%).
nonhydrogen atom;9 higher ligand efficiency defines a superior
compound. Interestingly, although methotrexate (3, MTX, Figure
1) is a much more potent compound (IC50 values against
ChDHFR and TgDHFR are 23 nM and 14 nM, respectively)
than TMP, it has the same ligand efficiency as TMP, implying
that the increased potency is largely dependent on the increased
molecular weight rather than a more optimal positioning of
pharmacophoric elements. It should be possible to design a
highly efficient ligand for DHFR that possesses the potency of
MTX while maintaining low molecular weight.
allows the chemist more confidence in the selection of com-
pounds slated for synthesis. Establishing reliable docking
procedures with known compounds allows more accurate
predictions of new compounds. In previous work,11 we docked
30 compounds into the crystal structure of ChDHFR and
established high (72.9%) correlations between the docking
scores and biological activity. Including protein flexibility in
the docking method and assessing an ensemble of the lowest-
energy docked complexes was critical to achieving this cor-
relation between calculated and experimental values.
Accurate computational prediction of protein-ligand interac-
tions is a great advantage in the ligand design process since it
In this manuscript, we present a novel, low molecular weight
scaffold for DHFR inhibitors that exhibits nanomolar potency

942 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5
Table 1. Inhibitory Potency of DHFR Ligands (IC50 Values in µM)
Pelphrey et al.
compound
R1
R2
ChDHFR
14
>1000
>1000
>1000
>1000
86.1 ( 7.3
10.2 ( 0.8
5.3 ( 0.1
0.169 ( 0.006
193.9 ( 10.5
141.2 ( 17.9
132.7 ( 13.1
5 ( 0.3
0.038 ( 0.0001
1.91 ( 0.1
TgDHFR
8
N/D
N/D
docking score vs ChDHFRa
TMP
7
8
9
10
14
16
20
21
24
25
27
28
37 (R)
38 (S)
N/A
N/A
N/A
N/A
N/A
H
CH3
H
CH3
H
CH3
H
CH3
CH3
CH3
N/A
N/A
N/A
N/A
N/A
H
5.94
-
-
-
N/D
90.2 ( 19.1
21.7 ( 3.3
0.88 ( 0.1
3.1 ( 0.4
-
6.98
7.36
7.50
-
-
-
-
-
H
CH3
CH3
OH
0.12 ( 0.0001
17.9 ( 2.6
9.1 ( 2.4
OH
OCH3
OCH3
CH3
CH3
9.8 ( 0.4
0.55 ( 0.03
0.0014 ( 0.00007
0.013 ( 0.001
7.84
8.37
a Docking scores are Surflex-Dock scores, in which higher numbers represent higher affinity. N/A: not applicable, N/D: not determined
against DHFR from both C. hominis and T. gondii. The
structures of the enzymes guided the design of derivative
compounds such that large increases in potency were achieved
during the synthesis of just 14 compounds from the parent, TMP.
While we have focused on increasing potency here, the lead-
like nature of these compounds allows the incorporation of
additional functionality for selectivity without overstepping
molecular weight boundaries.
(32 in Ch, 31 in Tg) forms an electrostatic interaction with the
protonated N1 and a hydrogen bond with the amino group at
position 2 as well as Thr (134 in Ch, 172 in Tg), Val (9 in Ch,
8 in Tg) and Phe (36 in Ch and 35 in Tg) that form van der
Waals interactions. The linker between the pteridine and the
para-aminobenzoic acid (pABA) is surrounded by one con-
served Thr (58 in Ch and 83 in Tg) and one nonconserved
residue: Cys 113 in Ch and Val 151 in Tg. The pABA ring is
bound in a less well-conserved hydrophobic pocket comprised
of Ile 62 in Ch (Met 87 in Tg), Leu 67 in Ch (Leu 94 in Tg)
and Leu 33 in Ch (Phe 32 in Tg).
Chemistry, Modeling, and Biological Evaluation
Structural Analysis of DHFR from C. hominis and T.
gondii. We have previously reported the crystal structure of
ChDHFR with different ligands;12,13 however, there is no
experimentally determined structure of TgDHFR. In previous
work11 we created a homology model of TgDHFR based on
the closely related structure of DHFR from Mus musculus (PDB
ID: 1U7014). The TgDHFR model was minimized, and a
Ramachandran analysis showed that the backbone geometry fell
outside allowed regions for only six residues in loop regions.
The model was further validated by docking eleven inhibitors
with determined IC50 values into the active site and achieving
a 50.2% correlation with the measured inhibition constants.
Superpositions of the crystal structure of ChDHFR and the
homology model of TgDHFR show that the two enzymes are
very similar, both in overall fold and in the identity and
arrangement of many active site residues (Figure 2a). Surround-
ing the pteridine ring system of methotrexate (Figure 2b), a
potent inhibitor modeled into both sites, are residues that are
conserved between the two species: an aspartic acid residue
In both species, models of TMP (2) show electrostatic and
hydrogen bond interactions between the protonated N1 and the
amino group at the C2 position of the pyrimidine ring and the
conserved Asp in the active site. This interaction between the
2,4-diaminopyrimidine and an acidic residue in the active site
is conserved across many species of DHFR.3,12,15-17 However,
the linker region of TMP appears to be too short to extend the
trimethoxyphenyl ring fully into the hydrophobic pocket nor-
mally occupied by the pABA ring of MTX, possibly explaining
the lower in Vitro potency of TMP against these species. Based
on these structural analyses, a relatively simple strategy emerged
to increase the potency of TMP for both of the target organisms
by extending the length of the linker between the two aromatic
rings.
Design and Synthesis of Extended TMP Analogues. In
order to test the hypothesis that extending the distance between
the diaminopyrimidine and phenyl rings of TMP would achieve
greater potency against ChDHFR and TgDHFR, we explored

Highly Efficient Ligands for DHFR
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5 943
Scheme 2a
Figure 3. (a) TMP modeled in the active site of ChDHFR. (b)
Compound 14 modeled in the active site of ChDHFR.
a Reagents and conditions: (a) LAH, Et2O, 0 °C, 1 h (93%); (b) Dess-
Martin periodinane, DCM, 25 °C, 1 h (96%); (c) dimethyl (1-diazo-2-
oxopropyl)phosphonate, K2CO3, MeOH, 0 °C, 1 h (52%); (d) 5, Pd(PPh3)2Cl2,
CuI, Et3N, DMF, 100 °C, 2 h (80%).
(increase from two rotatable bonds in TMP to three in 10). The
other three compounds in this series have lower internal entropy
but cannot reach conformations that are productive for binding.
Redesign in this series pointed toward analogues with a limited
number of rotatable bonds but more freedom for the two
aromatic rings to find appropriate binding pockets. The use of
a three-carbon propargylic tether appeared as a potential design,
as it maintains the same number of degrees of freedom as TMP
but allows the two rings to explore more structural space
independent of the other. Docking of the two alternative
propargyl-linked structures revealed that attaching the tri-
methoxyphenyl group to the methylene carbon rather than the
alkyne carbon would be preferred.
The homologated alkyne 13 required for the synthesis was
prepared in three operations from the commercially available
acid 11 (Scheme 2). Reduction of 11 to the alcohol, followed
by Dess-Martin oxidation, produced the corresponding alde-
hyde that was condensed with the Ohira-Bestmann20 reagent
to directly deliver the terminal acetylene 13. This alkyne was
engaged in Sonagashira coupling reactions with the previously
described 5 to give the parent compound 14. It was found that
14 inhibited ChDHFR with an IC50 value of 86 µM and
TgDHFR with an IC50 value of 22 µM (Table 1).
While the propargyl extended compound (14) did inhibit the
enzyme, its activity was unexpectedly lower than that of TMP.
Therefore we reexamined models of TMP and 14 in ChDHFR
to better understand this discrepancy (Figure 3). Extending the
phenyl ring did appear to increase interactions with the
hydrophobic pocket, specifically with Ile 62 and Leu 67, as
predicted. However, the extension of the phenyl also caused a
loss of lipophilic interactions with Leu 25, Leu 33, and Phe 36
and created a pocket near the C6 position of the pyrimidine
ring. This area is normally occupied by secondary ring fusions
in potent inhibitors such as MTX. Further analysis of the docked
TMP and 14 also revealed a second empty space near Cys 113
within the binding site.
A combination of Flo9821 and Surflex-Dock22,23 within the
Sybyl environment were used to assess docking scores for
methyl substitutions at the pyrimidine and propargyl locations
on 14. Flo98 performs a Monte Carlo search for the best docking
poses into a flexible target and saves an ensemble of protein:
ligand complexes with associated energy scores. Surflex-Dock
does not dock into the binding site but rather uses a protomol24
target that is created by probing the active site with small
the use of a two-carbon linker in place of the single methylene
bridge found in TMP. In previous work, the synthesis of six
5-ethynylpyrimidine derivatives with non-benzyl moieties had
been reported18 and these compounds appeared to inhibit T.
gondii and two species of fungal DHFR. TMP derivatives
containing saturated aliphatic (10), cis (8) and trans-olefinic
(9), or acetylenic (7) linkers were docked into the crystal
structure of ChDHFR, chosen because it was determined from
experimental data. TMP analogues with rigid olefinic and
acetylenic linkers allowed the diaminopyrimidine to dock into
the proper orientation but these linkers prevented the trimethox-
yphenyl ring from occupying the biologically relevant hydro-
phobic pocket. However, the high degree of flexibility allowed
by the saturated ethylene bridge appeared to allow both rings
to occupy their respective pockets. A direct and high-yielding
synthetic route to 10 (Scheme 1) was developed that incidentally
allowed for the preparation of this entire series of TMP
analogues.
The commercially available diaminopyrimidine 4 could be
directly iodinated at the C5 position to give 5. Direct conversion
to the tolan derivative 7 was achieved through a palladium-
catalyzed Sonagashira coupling reaction18,19 with trimethox-
yphenyl acetylene, 6, prepared by Corey-Fuchs20 extension of
the corresponding aldehyde. Catalytic hydrogenation of 7 was
next employed to prepare the fully saturated analogue 10.
Interestingly, under standard hydrogenation conditions, the
second reduction was found to be quite slow, and the cis-olefinic
linker 8 could be isolated and purified. Equilibration to the more
thermodynamically stable trans-isomer 9 was accomplished by
treatment with iodine and allowed easy access to the entire two-
carbon bridge series. Examination of these extended TMP
analogues in standard enzyme inhibition assays revealed that
all four analogues were inactive (Table 1). Poor activity was
anticipated for analogues 7, 8, and 9 from the docking studies,
but not for the fully saturated derivative 10.
Design and Synthesis of the Propargyl-Linked Series. The
failure to obtain active compounds in this series (7-10) was
presumed to be due to a mixture of entropic and conformational
effects. While compound 10 can easily adopt a conformation
that allows both aromatic rings to occupy their respective
binding pockets, it does so at a significant entropic penalty
induced by the organization around the highly flexible linker
Scheme 3

944 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5
Pelphrey et al.
Scheme 4a
a Reagents and conditions: (a) Ph3PdCHOMe, n-BuLi, THF, 0 °C, 1 h (73%); (b) concd HCl, THF, reflux, 3 h (93%); (c) CBr4, PPh3, DCM, 0 °C, 35
min (75%); (d) Mg, THF, reflux, 1.5 h (70%); (e) 5 or 15, Pd(PPh3)2Cl2, CuI, Et3N, DMF, 100 °C, 2 h; (f) ethynylmagnesium bromide, THF, 0 °C f 25
°C, 1 h (95%); (g) NaH, Me2SO4, THF, 25 °C, 1 h (90%)
Scheme 5a
a Reagents and conditions: (a) pivaloyl chloride, Et3N, oxazolidinone, n-BuLi, THF, -78 °C f 25 °C, 3 h; (b) LHMDS, MeI, THF, -78 °C f 0 °C,
2 h; (c) DIBAL-H, DCM, -78 °C, 2 h; (d) CBr4, PPh3, DCM, 0 °C, 35 min; (e) Mg, THF, reflux, 1 h; (f) 15, Pd(PPh3)2Cl2, CuI, Et3N, DMF, 100 °C, 2 h.
molecular fragments. Ligands are then fragmented and built into
the protomol based on an empirical scoring function25 that
accounts for hydrophobic, polar, repulsive, entropic, and sol-
vation terms.26
The docking score of 14 with individual methyl groups in
the two noted pockets was greater than that of the unsubstituted
compound 14, while the analogue with methyl groups in both
pockets simultaneously presented the highest score (Table 1).
Since Surflex-Dock cannot assess a racemic compound, both
the R and S enantiomers of the compound with a methyl at the
propargyl position were evaluated. Both appear equally viable
upon visual analysis and had many of the same interactions with
nearby residues. All complexes presented the ‘correct’ 2,4-
diaminopyrimidine positioning within the binding site, with
hydrogen bonds to Val 9, Val 10, Asp 32, and Thr 134.
The propargyl scaffold appeared to be superior to TMP for
placing substituents in these pockets as well as for synthetic
accessibility. The analogous substitutions on TMP have a large

Highly Efficient Ligands for DHFR
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5 945
Table 2. Ligand Efficiency of Selected Compounds
impact in the conformational distribution while interactions
between the C6 substitution and the aryl ring in the propargyl
series were computationally shown to not be significant. It
proved possible to extend the previous success of the Sonagash-
ira coupling reactions (Scheme 3) to install substitution at C6
of the extended inhibitors.
Cross-coupling of acetylene 13 with the known18 iodopyri-
midine 15 (prepared from commercially available 2-amino-4-
chloro-6-methylpyrimidine by amination and iodination) yielded
the C6-methyl derivative 16. As predicted by both visual
analysis of the docked compound and the docking scores (Table
1), compound 16 displayed improved potency over the unsub-
stituted analogue (14) against both enzymes with IC50 values
of 10.2 and 0.88 µM against ChDHFR and TgDHFR, respec-
tively (Table 1).
In previous work,4 we explored the second vacant pocket near
residue Cys 113 in ChDHFR by employing substitutions on the
methylene bridge of TMP. C7-methyl TMP (rac) has an IC50
value of 340 µM, and C7-ethyl TMP (rac) showed moderate
improvement over TMP, with an IC50 value of 4 µM. Again,
modifying the propargyl extended scaffold of 14 and 16
appeared to be a better route for exploiting this pocket, as the
substitution would potentially form higher affinity interactions
with Cys 113. We examined a small series of three different
substituents (methyl, hydroxyl, and methoxy) at the propargylic
position in an attempt to increase the potency of compounds
14 and 16. The new terminal acetylenes for these derivatives
were prepared, as racemates, in a straightforward manner
(Scheme 4).
compound
ChDHFR
TgDHFR
TMP
MTX
14
16
37
0.32
0.32
0.24
0.29
0.41
0.33
0.33
0.28
0.35
0.49
15 gave the two enantiomers 37 and 38 in 90 and 95% ee,
respectively as determined by HPLC analysis using a chiral
stationary phase column.
The two enantiomers exhibited quite different affinities for
the protozoal enzymes with the R enantiomer inhibiting
ChDHFR with an IC50 value of 38 nM and TgDHFR with an
IC50 value of 1.4 nM. The opposite enantiomer, S, inhibited
ChDHFR and TgDHFR with IC50 values of 1.9 µM and 13 nM,
respectively.
In addition to an improvement in potency, this series of
compounds shows an improvement in ligand efficiency (Table
2). The ligand efficiency of 14 decreased relative to TMP since
the additional molecular weight due to the acetylene did not
lead to appropriate increases in potency. However, the methyl
group additions to the propargyl scaffold exhibited large
increases in ligand efficiency. In the ChDHFR case, the C6 and
propargyl methyl groups yielded 16% and 43% increases in
ligand efficiency, respectively, and for TgDHFR, the C6 and
propargyl methyl groups yielded 26% and 44% increases in
ligand efficiency, respectively.
Discussion
The commercially available trimethoxyacetophenone 17 was
homologated to the aldehyde 18 through Wittig condensation
and hydrolysis of the resulting enol ether.27 A modified Corey-
Fuchs homologation28 provided the racemic acetylene 19 that
was coupled with both iodopyrimidines 5 and 15 to produce
the corresponding analogues 20 and 21. Installation of a
hydroxyl or methoxy substituent could be accomplished in a
straightforward manner from aldehyde 22. This was converted
to the propargyl alcohol 23 through the addition of acetylide,
and then standard Sonagashira couplings on 23 gave analogues
24 and 25 in high yield. Alternatively, the alcohol could be
converted to the corresponding methyl ether 26 under standard
conditions and cross-coupled in an analogous manner to deliver
pyrimidines 27 and 28. Again, as predicted from the docked
complexes, the methyl-substituted propargyl compound, 21,
exhibited marked improvement in potency with IC50 values of
169 nM and 120 nM against ChDHFR and TgDHFR, respec-
tively. The hydroxy- and methoxy-substituted propargyl com-
pounds did not exhibit the same improvement (Table 1).
With the excellent performance of the doubly methylated
derivative 21 in ChDHFR and TgDHFR inhibition assays, it
was logical to examine the individual isomers at the stereogenic
center to determine if one of the enantiomers was more active
than the other. To accomplish the synthesis of the two
enantiomeric analogues, we employed an Evans oxazolidinone-
mediated asymmetric alkylation29 (Scheme 5).
Addition of two different lithio oxazolidinones to the mixed
anhydride derived from acid 11 led to imides 29 and 30 that
could be alkylated to produce 31 and 32 with fairly high
diastereoselectivity. Unfortunately, the minor diastereomer could
not be removed during purification and was taken on in the
following steps. Partial reduction of the imide with DIBAL-H
produced the aldehydes 33 and 34 that were immediately
homologated as described in the racemic synthesis to give
terminal alkynes 35 and 36. Cross-coupling with iodopyrimidine
Through analysis of the crystal structure of ChDHFR and a
validated homology model of TgDHFR, it was noted that
structural features common to both enzymes may contribute to
the poor activity of the archetypical inhibitor, trimethoprim.
Specifically, the trimethoxyphenyl ring did not appear to extend
deep enough into the hydrophobic pocket which is normally
occupied by the natural substrate, dihydrofolate. Guided by this
analysis, we modified the TMP structure to arrive at highly
potent inhibitors (368-fold improvement for ChDHFR and 5714-
fold for TgDHFR, over the parent TMP) through the synthesis
of just 14 new analogues.
The models of 37 bound to ChDHFR and TgDHFR (Figure
4) nicely illustrate why this inhibitor is highly potent. In
ChDHFR (Figure 4a), the C6 methyl group forms ideal van
der Waals interactions with Phe 36, Leu 33, and Leu 25. The
propargyl methyl group forms van der Waals interactions with
Cys 113 and, remotely, with Ile 62. In TgDHFR (Figure 4b),
the C6 methyl group forms van der Waals interactions with Phe
35 and Phe 32; the propargyl methyl group forms van der Waals
interactions with Met 87 and Val 151.
Each enantiomer of the stereogenic propargyl position,
compounds 37 and 38, possesses higher affinity than compounds
that are unsubstituted at that position (14 or 16). In fact, this
increase in activity for each enantiomer was predicted compu-
tationally (Table 1). The fact that the S enantiomer was predicted
to be more active than the R, when the biological evaluation
shows that the R is more active, is a limitation of the
computational analysis. Interestingly, in the T. gondii case, each
enantiomer is more active in the enzyme assay than the racemic
mixture (21). Although this appears counterintuitive, it is most
likely due to the complex kinetic competition between substrate
and inhibitors in the assay.
The work in this manuscript represents the first efforts in a
two-tiered approach to achieving both potent and selective

946 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5
Pelphrey et al.
Figure 4. Models of 37 bound to (a) ChDHFR and (b) TgDHFR. In both cases, the inhibitor is colored blue and the protein is shown in salmon.
Key residues in the active site are numbered.
at least four times. IC50 values and their standard deviations were
calculated in the presence of varying concentrations of inhibitor
near the IC50 concentration.
inhibitors of DHFR from these parasitic protozoa. The low
molecular weight of the lead compound in this work allows
the installation of additional functionality to achieve selectivity
against the human form of DHFR. Structural comparisons of
ChDHFR and hDHFR12 reveal that ChDHFR lacks a loop at
the active site that is present in hDHFR. In future work, we
plan to explore derivatives of the compounds reported here that
extend functionality toward the loop, thus preventing binding
to hDHFR and increasing selectivity. For TgDHFR, there are
residue substitutions in the active site relative to hDHFR,
including Val 5 and Met 87. Derivative compounds that exploit
these differences will be used to increase selectivity.
Several heterocycles have been reported as potent inhibitors
of DHFR from T. gondii, Pneumocystis carinii, and Mycobac-
terium aVium. Some of these, including a 2,4-diamino-5-(2′,5′-
disubstituted benzyl)pyrimidine,30 a 2,4-diamino-6-[2′-ω-car-
boxyalkyl)dibenzazepine]methylpteridine,31 and three 2,4-
diamino-5-deazapteridines32,33 display nanomolar potency against
TgDHFR. Like MTX, the increased potency of several of these
previously reported inhibitors is related to the use of secondary
ring fusions to the diaminopyrimidine core. On the basis of our
analysis, these additional rings provide increased van der Waal
contacts and also project functionality deeper into the hydro-
phobic pocket, much like the C6 substituent and the propargyl
linker in our design, respectively. The acetylene moiety provides
the rigidification and spacing in this system that is attributed to
the second ring fusion in these other inhibitors but does so with
two key additional benefits. First, there is only a slight increase
in molecular weight relative to TMP, and in fact the best
inhibitor presented here has an increase of only 52 Da in mass
to achieve an increased potency of 368-fold for ChDHFR and
5714-fold for TgDHFR. Second, this new scaffold has much
greater synthetic accessibility, owing in large part to the key
Sonagashira coupling which allows for a convergent and
modular approach. This type of synthetic strategy should be
easily amenable to more extensive analogue development and
even parallel synthetic strategies. The key features of these new
TMP analogues make them excellent candidates for further
development.
TgDHFR: DHFR cloned from T. gondii DHFR-TS was
determined to be insoluble after refolding at relatively low protein
concentrations (>1 mg/mL). The presence of an unstructured loop
region (residues 43-70) that is not present in other DHFR proteins
was predicted to render the protein insoluble. These residues were
removed, leaving a five residue loop region characteristic of several
other species of DHFR. The removal of this unstructured loop
region yielded a soluble preparation of TgDHFR at concentrations
in excess of 18 mg/mL, a single species by native gel electrophoresis
and with an activity level equal to T. gondii DHFR-TS and
characteristic of other DHFR proteins after refolding. See Support-
ing Information for cloning and purification details. Activity assays
followed the same procedure as those used for ChDHFR-TS.
Calculation of Ligand Efficiency. Ligand efficiency is expressed
as binding energy per non-hydrogen atom (∆G/Nnon-hydrogen
)
atoms
where ∆G ) -RT ln Kd, although IC50 values can be substituted
for Kd.9 Therefore, ligand efficiencies were calculated using the
quotient of ∆G ) (1.99 cal K-1 mol-1)(300 K) (ln IC50)(0.001 kcal/
cal) and the number of non-hydrogen atoms.
Computational Modeling. All ligands were created in Sybyl
(Tripos Inc.) and checked for correct geometries. These were
selectively protonated at N1 of the 2,4-diaminopyrimidine ring.
Formal charges were calculated automatically.
Chain A of the 1SEJ crystal structure12 was selected as the
ChDHFR receptor. The site was prepared by removing all waters,
adding hydrogens, checking for steric clashes, and calculating
formal charges. The homology model of TgDHFR was created as
described in the text and with further details.11
For docking procedures using Flo98, the active sites of both
proteins were defined as all residues with an atom falling within
an 11 Å sphere around the cocrystallized ligand. The automatic
protomol mapping protocol within Surflex-Dock explored the entire
active site and was used for docking purposes. The cocrystallized
NADPH was included in the definition of the active site.
Libraries of the analogues were docked using Flo98 as discussed
previously11 and also using Surflex-Dock as a Sybyl module. All
docking results were checked for correct orientation as defined by
the conserved hydrogen bond interactions between the protonated
N1 of the 2,4-diaminopyrimidine and Asp 32. The resulting
conformational energies were also checked for steric clashes and
unrealistic geometries.
HPLC. Samples (0.1 mg/mL in 14% CHCl3, 10 mM Na2HPO4,
pH 6) were analyzed isocratically (4% CH3CN, 10 mM Na2HPO4
pH 6, 1.5 mL/min) using a Chrom Tech Chiral-AGP column (4.0
mm × 100 mm) and a Beckman Coulter System Gold HPLC with
PDA detector.
General. The 1H and 13C spectra were recorded at 500 and 125
MHz, respectively, on a Varian Inova 500 spectrometer. All melting
points are uncorrected. High-resolution mass spectrometry was
performed on a Kratos MS-50 spectrometer by the Washington
University Mass Spectrometry laboratory. Elemental analyses were
provided by Atlantic Microlabs, Inc. All reagents were used directly
from commercial sources unless otherwise stated.
Experimental Section
Enzyme Expression, Purification, and Assays. ChDHFR:
ChDHFR-TS was expressed in E. coli and purified using a
methotrexate agarose column (Sigma).13 Enzyme activity assays
were performed by monitoring the change in UV absorbance at
340 nm in a solution containing 50 mM tris(hydroxymethyl)methyl-
2-aminoethanesulfonic acid pH 7.0, 1 mM EDTA, 75 µM 2-mer-
captoethanol, 1 mg/mL bovine serum albumin, 1 mM dihydrofolate
(Eprova), and 100 µM nicotinamide adenine dinucleotide phosphate
(NADPH) (Sigma) and limiting concentrations of enzyme. Enzyme
and inhibitor were allowed to incubate for 5 min before adding
dihydrofolate to initiate the reaction. Enzyme assays were performed

Highly Efficient Ligands for DHFR
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5 947
General Procedure for Sonagashira Couplings. To a sealed
tube was added the iodopyrimidine (1.0 mmol). DMF (5.0 mL)
was added and the mixture stirred until all solids dissolved. Pd-
(PPh3)2Cl2 (49.0 mg, 0.07 mmol) was added followed by CuI (13.3
mg, 0.07 mmol), Et3N (5.0 mL), and the aryl acetylene (2.0 mmol).
The tube was sealed and placed into a 100 °C oil bath. The reaction
was stirred at 100 °C for 2 h and cooled to room temperature. The
solvent was removed and the residue purified by flash chromatog-
raphy (SiO2, 30 g) using 2% MeOH in CHCl3 as the eluent to afford
the coupled products. Analytical samples were prepared by recrys-
tallization from MeCN.
156.2, 154.3, 138.6, 134.0, 133.0, 122.6, 107.3, 106.4, 61.1, 56.4;
HRESI [M + H] 303.1446 (calculated C15H19N4O3: 303.1457).
(E)-2,4-Diamino-5-(3,4,5-trimethoxystyryl)pyrimidine (9). 8
(45.0 mg, 0.149 mmol) was suspended in dry THF (1.5 mL). I2
(4.0 mg, 0.0149 mmol) was added and the reaction allowed to stir
at 23 °C for 30 min after which time all material was in solution.
The red solution was diluted with THF (5.0 mL), and saturated
Na2S2O3 (1.0 mL) was added. The layers were separated and the
organics washed with water (5.0 mL) and brine (5.0 mL) and dried
over anhydrous MgSO4. The solvent was removed and the residue
purified by flash chromatography (SiO2, 5.0 g) using 5% MeOH
in CHCl3 as the eluent to afford 9 as a yellow powder (36 mg,
80%) which was recrystallized from MeCN: Rf ) 0.28 (9:1, CHCl3:
MeOH) ; mp ) 182-184 °C; 1H NMR (acetone-d6) δ 8.09 (s, 1H),
7.07 (d, J ) 15.9 Hz, 1H), 6.85 (s, 2H), 6.82 (d, J ) 15.9 Hz, 1H),
5.94 (s, 2H), 5.51 (s, 2H), 3.83 (s, 6H), 3.71 (s, 3H); 13C NMR
(CDCl3) δ 161.6, 161.3, 154.7, 153.4, 138.0, 132.9, 129.1, 119.8,
107.0, 103.4, 60.9, 56.1; HRESI [M + H] 303.1446 (calculated
C15H19N4O3: 303.1457).
2,4-Diamino-5-iodopyrimidine (5). To a flame-dried 100 mL
round-bottom flask was added 2,4-diaminopyrimidine (1.0 g, 9.08
mmol) in MeOH (30 mL) followed by dropwise addition of ICl
(30 mL, 29.06 mmol). The solution was stirred at 25 °C for 15 h
and then the solvent removed under reduced pressure. The resulting
viscous oil was stirred in Et2O (40 mL) for 45 min. The resulting
solid was filtered off and washed with Et2O (3 × 10 mL) to afford
the HCl salt as a yellow solid (3.14 g). The crude salt was suspended
in 1.0 N NaOH (100 mL) and stirred at 25 °C for 2 h. The solids
were filtered, washed with water (2 × 10 mL), and dried to afford
5 as a brown powder (1.71 g, 80%). An analytical sample was
prepared by recrystallization from MeCN to give 5 as colorless
2,4-Diamino-5-(3,4,5-trimethoxyphenethyl)pyrimidine (10). 7
(100 mg, 0.333 mmol) was placed into a 100 mL shaker flask.
MeOH (20 mL) was added, and the mixture was swirled until the
solids dissolved. 10% Pd/C (100 mg) was added, and the suspension
was placed into a hydrogenation apparatus. The reaction was
allowed to run for 10 h at 50 psi H2. The solid residue was filtered
through Celite and washed with methanol (10 mL). The solvent
was removed, and the residue was purified by flash chromatography
(SiO2, 10 g) using 10% MeOH in CHCl3 as the eluent to afford 10
as a white powder (88.0 mg, 87%): Rf ) 0.06 (9:1, CHCl3:MeOH);
1
crystals: Rf ) 0.25 (9:1, CHCl3:MeOH); mp ) 212-214 °C; H
NMR (DMSO-d6) δ 7.92 (s, 1H), 6.40 (s, 2H), 6.10 (s, 2H); 13C
NMR (DMSO-d6) δ 162.8, 162.7, 162.0, 61.2; HREI[M+] 235.9559
(calculated C4H5IN4: 235.9559); Anal. (C4H5IN4) C, H, N.
5-Ethynyl-1,2,3-trimethoxybenzene (6). To a flame-dried 50
mL round-bottom flask was added CBr4 (2.54 g, 7.65 mmol). DCM
(20 mL) was added and the solution cooled to 0 °C. Ph3P (4.01 g,
15.30 mmol) was added and the solution stirred at 0 °C for 15
min. 3,4,5-Trimethoxybenzaldehyde (1.0 g, 5.10 mmol) in DCM
(6.0 mL) was added dropwise. The solution was stirred at 0 °C for
5 min. The solvent was removed under reduced pressure, and the
resulting oil was filtered through a plug of silica and washed with
Hex:EtOAc (9:1, 500 mL; 4:1, 500 mL). The combined organics
were concentrated to give the crude dibromo-olefin (2.31 g, 128%),
which was dissolved in THF (60 mL) and cooled to -78 °C. n-BuLi
(13.1 mL, 19.68 mmol, 1.5 M) was added dropwise and the solution
stirred at -78 °C for 30 min. Saturated NH4Cl (10 mL) was added
and the solution warmed to room temperature. The layers were
separated, and the organics were washed with brine (10 mL), dried
over anhydrous Na2SO4, and concentrated. The residue was purified
by flash chromatography (SiO2, 80 g) using 10% EtOAc in hexanes
as the eluent to afford 6 as a colorless oil (0.770 g, 79%): Rf )
0.30 (4:1, Hex:EtOAc); 1H NMR (CDCl3) δ 6.71 (s, 2H), 3.84 (s,
3H), 3.83 (s, 6H), 3.03 (s, 1H); 13C NMR (CDCl3) δ 153.1, 117.1,
109.4, 103.3, 83.8, 76.4, 61.0, 56.2, 56.1; HRFAB [M + Li]
199.0952 (calculated C11H12O3Li: 199.0946).
1
mp ) 155-157 °C; H NMR (DMSO-d6) δ 7.46 (s, 1H), 6.56 (s,
2H), 6.22 (s, 2H), 5.65 (s, 2H), 3.74 (s, 6H), 3.61 (s, 3H), 2.65-
2.62 (m, 2H), 2.52-2.49 (m, 2H); 13C NMR (DMSO-d6) δ 162.3,
162.0, 155.0, 152.6, 137.5, 135.5, 105.8, 105.8, 60.0, 55.6, 35.0,
29.1; Anal. (C15H20N4O3) C, H, N.
1,2,3-Trimethoxy-5-(prop-2-ynyl)benzene (13). To a flame-
dried 1 L flask was added LiAlH4 (3.85 g, 101.4 mmol). Et2O (250
mL) was added and the suspension cooled to 0 °C. Trimethox-
yphenylacetic acid (15.3 g, 67.6 mmol) in Et2O (50 mL) was added
dropwise and the reaction allowed to stir at 0 °C for 1 h. Water
(50 mL) was added and the solution warmed to room temperature.
The solids were filtered off and washed with Et2O (3 × 25 mL).
The combined organics were washed with brine (25 mL), dried
over anhydrous Na2SO4, and concentrated to give the crude alcohol
(13.4 g, 93%). The crude alcohol (13.3 g, 62.7 mmol) was dissolved
in DCM (125 mL). Dess-Martin periodinane (39.8 g, 94.0 mmol)
was added and the reaction stirred at room temperature for 1 h.
Saturated NaHCO3:saturated Na2S2O3 (1:1, v:v, 20 mL) was added
and the reaction stirred for 30 min. The organic layer was washed
with brine (25 mL), dried over anhydrous Na2SO4, and concentrated
to give aldehyde 12 (13.1 g, 96%). Spectra were identical to
literature values.34
2,4-Diamino-5-(2-(3,4,5-trimethoxyphenyl)ethynyl)pyrimi-
dine (7). 5 (236 mg) was allowed to react with 6 (384 mg) as per
the general procedure to afford 7 as a white powder (270 mg,
The Ohira-Bestmann reagent (15.1 g, 78.6 mmol) was dissolved
in MeOH (260 mL) and the reaction cooled to 0 °C. 12 (11.0 g,
52.3 mmol) was added followed by K2CO3 (15.9 g, 115.0 mmol).
The reaction was allowed to stir at 0 °C for 1 h and warmed to
room temperature, and the solids were filtered. The organics were
concentrated and the residue purified by flash chromatography
(SiO2, 500 g) using 10% EtOAc in hexanes as the eluent to afford
1
90%): Rf ) 0.28 (9:1, CHCl3:MeOH); mp ) 202-204 °C; H
NMR (DMSO-d6) δ 7.94 (s, 1H), 6.90 (s, 2H), 6.37 (s, 2H), 3.79
(s, 6H), 3.67 (s, 3H); 13C NMR (DMSO-d6) δ 163.4, 162.2, 159.4,
159.2, 152.8, 137.6, 118.6, 108.3, 108.2, 94.7, 89.8, 83.4, 60.2,
60.1, 56.0, 55.9; Anal. (C15H16N4O3) C, H, N.
(Z)-2,4-Diamino-5-(3,4,5-trimethoxystyryl)pyrimidine (8). 7
(300 mg, 1.00 mmol) was placed into a 100 mL shaker flask. EtOH
(30 mL) was added, and the mixture was swirled until all solids
dissolved. 5% Pd/C (50 mg) was added, and the suspension was
placed into a hydrogenation apparatus. The reaction was allowed
to run at 45 psi H2 for 5 h. The mixture was filtered through Celite
and washed with EtOAc (15 mL). The solvent was removed and
the residue purified by flash chromatography (SiO2, 30 g) using
5% MeOH in CHCl3 as the eluent to afford 8 as a white powder
(100 mg, 33%, 81% borsm) which was recrystallized from
1
13 as a colorless oil (5.61 g, 52%): Rf ) 0.20 (10%EtOAc); H
NMR (CDCl3) δ 6.58 (s, 2 H), 3.86 (s, 6 H), 3.83 (s, 3H), 3.55 (s,
2 H), 2.22 (s, 1 H); 13C (CDCl3) δ 153.28, 136.75, 131.70, 104.89,
81.87, 70.69, 60.80, 56.06, 24.99; HREI[M+] 206.0943 (calculated
C12H14O3: 206.0943).
2,4-Diamino-5-(3-(3,4,5-trimethoxyphenyl)prop-1-ynyl)pyri-
midine (14). 5 (236 mg) was allowed to react with 13 (412 mg) as
per the general procedure to afford 14 as a white powder (251 mg,
80%): Rf ) 0.23 (9:1, CHCl3:MeOH); mp ) decomposed above
190 °C; 1H NMR (DMSO-d6) δ 7.84 (s, 1H), 6.71 (s, 2H), 6.26 (s,
2H), 3.80 (s, 2H), 3.77 (s, 6H), 3.63 (s, 3H); 13C NMR (DMSO-
d6) δ 163.8, 162.2, 158.7 152.8, 136.0, 132.8, 105.2, 93.1, 90.1
76.5, 60.1, 55.8, 25.6; Anal. (C16H18N4O3) C, H, N.
1
MeCN: Rf ) 0.33 (9:1, CHCl3:MeOH) ; mp ) 175-177 °C; H
NMR (acetone-d6) δ 7.69 (s, 1H), 6.63 (s, 2H), 6.47 (d, J ) 11.9
Hz, 1H), 6.27 (d, J ) 11.9 Hz, 1H), 5.66 (s, 2H), 5.49 (s, 2H),
3.68 (s, 3H), 3.67 (s, 6H); 13C NMR (CD3OD) δ 163.5, 163.3,

948 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5
Pelphrey et al.
1
2,4-Diamino-5-iodo-6-methylpyrimidine (15). To a 100 mL
steel pressure vessel was added 2-amino-4-chloro-6-methylpyri-
midine (4.00 g, 28.00 mmol). A saturated solution of NH3 in MeOH
(40 mL) was added and the vessel sealed. The pressure vessel was
placed into a 160 °C oil bath and heated for 15 h. The vessel was
cooled to 0 °C and opened and the solvent removed to give the
crude diaminopyrimidine (4.4 g) as the HCl salt. The crude salt
was dissolved in MeOH (82.5 mL), and ICl (82.5 mL) was added
dropwise over 50 min. The reaction was stirred at 25 °C for 14 h
and the solvent removed. The viscous oil was stirred in Et2O (300
mL) for 30 min. The resulting solid was filtered and washed with
Et2O (3 × 20 mL) to give the crude iodinated pyrimidine as a
yellow solid (10.1 g). The solids were suspended in 1.0 N NaOH
(300 mL) and stirred at 25 °C for 2 h. The solids were filtered,
washed with water (2 × 20 mL), and allowed to dry to afford 15
as a white powder (5.6 g, 80%). An analytical sample was prepared
by recrystallization from MeCN to give 15 as colorless crystals:
Rf ) 0.25 (9:1, CHCl3:MeOH); mp ) 154-156 °C; 1H NMR
(DMSO-d6) δ 6.38 (s, 2H), 6.09 (s, 2H), 2.24 (s, 3H); 13C NMR
(DMSO-d6) δ 166.7, 163.0, 162.3, 63.8, 28.4; HREI[M+] 249.9715
(calculated C5H7IN4: 249.9715); Anal. (C5H7IN4) C, H, N.
2,4-Diamino-5-(3-(3,4,5-trimethoxyphenyl)prop-1-ynyl)-6-me-
thylpyrimidine (16). 15 (250 mg) was allowed to react with 13
(412 mg) as per the general procedure to afford 16 as a white
powder (280 mg, 85%): Rf ) 0.31 (9:1, CHCl3:MeOH); mp )
70%): Rf ) 0.29 (4:1, Hex:EtOAc); H NMR (CDCl3) δ 6.62 (s,
2H), 3.88 (s, 6H), 3.84 (s, 3H), 3.74-3.69 (m, 1H), 2.29 (d, J )
2.7 Hz, 1H), 1.52 (d, J ) 7.1 Hz, 3H); 13C NMR (CDCl3) δ 153.4,
138.5, 136.9, 104.0, 104.0, 87.2, 70.5, 61.0, 56.3, 32.1, 24.5;
HRFAB [M + Li] 227.1243 (calculated C13H16O3Li: 227.1259).
2,4-Diamino-5-(3-(3,4,5-trimethoxyphenyl)but-1-ynyl)pyrimi-
dine (20). 5 (236 mg) was allowed to react with 19 (440 mg) as
per the general procedure to afford 20 as a yellow powder (295
mg, 90%): Rf ) 0.29 (9:1, CHCl3:MeOH); mp ) 220-222 °C;
1H NMR (DMSO-d6) δ 7.85 (s, 1H), 6.75 (s, 2H), 6.28 (s, 2H),
3.99 (q, J ) 6.8 Hz, 1H), 3.78 (s, 6H), 3.63 (s, 3H), 1.50 (d, J )
7.1 Hz, 3H); 13C NMR (DMSO-d6) δ 163.7, 152.8, 139.3, 136.0,
104.2, 104.1, 97.9, 76.2, 60.0, 60.0, 55.9, 55.8, 32.3, 24.3; Anal.
(C17H20N4O3) C, H, N.
2,4-Diamino-5-(3-(3,4,5-trimethoxyphenyl)but-1-ynyl)-6-me-
thylpyrimidine (21). 15 (250 mg) was allowed to react with 19
(440 mg) as per the general procedure to afford 21 as a white
powder (294 mg, 86%): Rf ) 0.29 (9:1, CHCl3:MeOH); mp )
1
191-193 °C; H NMR (DMSO-d6) δ 6.76 (s, 2H), 6.19 (s, 2H),
4.02 (q, J ) 7.1 Hz, 1H), 3.77 (s, 6H), 3.63 (s, 3H), 2.21 (s, 3H),
1.51 (d, J ) 7.1 Hz, 3H); 13C NMR (DMSO-d6) δ 167.0, 164.1,
161.0, 152.8, 139.4, 136.0, 104.0, 100.8, 88.6, 76.4, 60.0, 55.8,
32.5, 24.6, 22.5; Anal. (C18H22N4O3) C, H, N.
1-(3,4,5-Trimethoxyphenyl)prop-2-yn-1-ol (23). To a flame-
dried 250 mL round-bottom flask was added 3,4,5-trimethoxyben-
zaldehyde (3.92 g, 20.0 mmol). THF (40 mL) was added and the
solution was cooled to 0 °C. Ethynylmagnesium bromide (48.0 mL,
24.0 mmol, 0.5 M) was added dropwise. The solution was stirred
at 0 °C for 30 min, warmed to 25 °C, and stirred for 30 min.
Saturated NH4Cl (5.0 mL) was added, and the layers were separated.
The aqueous layer was extracted with Et2O (3 × 5 mL). The
combined organics were washed with brine (10 mL), dried over
anhydrous Na2SO4, and concentrated. The residue was purified by
flash chromatography (SiO2, 100 g) using 20% EtOAc in hexanes
as the eluent to afford 23 as a yellow oil (4.22 g, 95%): Rf ) 0.25
(1:1, Hex:EtOAc); 1H NMR (CDCl3) δ 6.77 (s, 2H), 5.39 (dd, J )
5.5, 2.0 Hz, 1H), 3.86 (s, 6H), 3.82 (s, 3H), 2.76 (d, J ) 5.9 Hz,
1H); 13C NMR (CDCl3) δ 153.4, 138.1, 135.9, 103.7, 83.6, 74.9,
64.5, 61.0, 56.2; HRFAB [M + Li] 229.1054 (calculated C12H14O4-
Li: 229.1052).
1
164-166 °C; H NMR (DMSO-d6) δ 6.72 (s, 2H), 6.42 (s, 2H),
3.85 (s, 2H), 3.76 (s, 6H), 3.63 (s, 3H), 2.24 (s, 3H); 13C NMR
(DMSO-d6) δ 165.0, 164.3, 159.7, 152.8, 135.9, 132.8, 105.0, 96.3,
89.3, 75.8, 60.0, 55.8, 25.7, 21.7; HRFAB [M + Li] 335.1679
(calculated C17H20N4O3Li: 335.1695).
5-(But-3-yn-2-yl)-1,2,3-trimethoxybenzene (19). To a flame-
dried 250 mL round-bottom flask was added methoxymethyltriph-
enylphosphonium bromide (10.28 g, 30.0 mmol). THF (100 mL)
was added and the solution cooled to 0 °C. n-BuLi (14.0 mL, 30.0
mmol, 2.2 M) was added dropwise and the solution stirred at 0 °C
for 30 min. 3,4,5-Trimethoxyacetophenone (5.26 g, 25.0 mmol) in
THF (25 mL) was added dropwise. The reaction was stirred at 0
°C for 30 min, and then water (30 mL) was added. The layers were
separated, and the aqueous layer was extracted with Et2O (3 × 20
mL). The combined organics were washed with brine (30 mL),
dried over anhydrous Na2SO4, and concentrated. The residue was
purified by flash chromatography (SiO2, 50 g) using 5% EtOAc in
hexanes as eluent to afford the enol ether (4.35 g, 73%).
The enol ether (4.35 g, 18.26 mmol) was dissolved in THF (37.0
mL). Concentrated HCl (3.0 mL) was added and the solution heated
to reflux. The solution was stirred at reflux for 3 h and allowed to
cool to room temperature. Water (10 mL) was added, and the
organics were washed with sat. NaHCO3 (10 mL) and brine (10
mL), dried over anhydrous Na2SO4, and concentrated. The residue
was purified by flash chromatography (SiO2, 50 g) using 20%
EtOAc in hexanes as the eluent to afford 18 (3.82 g, 93%). Spectra
were the identical to literature values.27
CBr4 (8.47 g, 25.55 mmol) was dissolved in DCM (150 mL)
and cooled to 0 °C. Ph3P (13.4 g, 51.09 mmol) was added and the
solution stirred at 0 °C for 5 min. 18 (3.82 g, 17.03 mmol) in DCM
(20 mL) was added dropwise. The reaction was stirred at 0 °C for
30 min and then poured into ice-cooled Et2O (500 mL). The solids
were filtered through Celite and washed with Et2O (3 × 50 mL).
The combined organics were concentrated. The residue was filtered
through a plug of silica and washed with hexanes (100 mL) followed
by 10% EtOAc in hexanes (5 × 100 mL). The combined organics
were concentrated to give the intermediate dibromide (4.86 g, 75%)
which was used directly for the next reaction.
Mg (0.621 g, 25.57 mmol) was suspended in THF (2.0 mL).
1,2-Dibromoethane (0.442 mL, 5.12 mmol) was added and the
reaction stirred at 25 °C for 30 min. The dibromide (4.86 g, 12.79
mmol) in THF (11.0 mL) was added dropwise and the solution
heated to reflux where it was stirred for 1 h. The solution was cooled
to room temperature and the solvent removed. The residue was
purified by flash chromatography (SiO2, 150 g) using 10% EtOAc
in hexanes as the eluent to afford 19 as a colorless oil (1.97 g,
3-(2,4-Diaminopyrimidin-5-yl)-1-(3,4,5-trimethoxyphenyl)-
prop-2-yn-1-ol (24). 5 (236 mg) was allowed to react with 23 (444
mg) as per the general procedure to afford 24 as a yellow powder
(257 mg, 78%): Rf ) 0.12 (9:1, CHCl3:MeOH); mp ) 203-205
1
°C; H NMR (DMSO-d6) δ 7.84 (s, 1H), 6.83 (s, 2H), 6.34 (s,
2H), 6.01 (d, J ) 5.6 Hz, 1H), 5.50 (d, J ) 5.4 Hz, 1H), 3.78 (s,
6H), 3.65 (s, 3H); 13C NMR (DMSO-d6) δ 163.8, 162.3, 158.2,
152.7, 138.3, 136.7, 109.3, 103.6, 96.3, 79.3, 63.4, 60.0, 55.8; Anal.
(C16H18N4O4) C, H, N.
3-(2,4-Diamino-6-methylpyrimidin-5-yl)-1-(3,4,5-trimethox-
yphenyl)prop-2-yn-1-ol (25). 15 (250 mg) was allowed to react
with 23 (444 mg) as per the general procedure to afford 25 as a
yellow powder (275 mg, 80%): Rf ) 0.12 (9:1, CHCl3:MeOH);
1
mp ) 174-176 °C; H NMR (DMSO-d6) δ 6.85 (s, 2H), 6.27 (s,
2H), 5.98 (d, J ) 5.6 Hz, 1H), 5.53 (d, J ) 5.4 Hz, 1H), 3.78 (s,
6H), 3.65 (s, 3H), 2.19 (s, 3H), 1.51 (d, J ) 7.1 Hz, 3H); 13C NMR
(DMSO-d6) δ 167.1, 164.3, 161.2, 152.7, 138.3, 136.7, 103.7, 99.1,
87.9, 79.5, 63.6, 60.0, 55.8, 22.5; HRFAB [M + Li] 351.1638
(calculated C17H20N4O4Li: 351.1645).
1,2,3-Trimethoxy-5-(1-methoxyprop-2-ynyl)benzene (26). To
a flame-dried 100 mL round-bottom flask was added NaH (0.240
mg, 6.0 mmol) that had been prewashed with pentane (3 × 15 mL)
and dried. THF (48 mL) was added and the suspension cooled to
0 °C. 23 (1.11 g, 5.0 mmol) in THF (2.0 mL) was added dropwise
and the reaction stirred at 0 °C for 25 min. Me2SO4 (0.571 mL,
6.0 mmol) was added and the reaction stirred at 0 °C for 20 min.
Water (10 mL) was added, and the layers were separated. The
aqueous layer was extracted with Et2O (20 mL). The combined
organics were washed with brine (20 mL), dried over anhydrous
Na2SO4, and concentrated. The residue was purified by flash
chromatography (SiO2, 20 g) using 10% EtOAc in hexanes as the

Highly Efficient Ligands for DHFR
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5 949
eluent to afford 26 as a colorless oil (1.06 g, 90%): Rf ) 0.18
layer was extracted with EtOAc (2 × 10 mL). The combined
organics were washed with brine (10 mL), dried over anhydrous
Na2SO4, and concentrated. The residue was purified by flash
chromatography (SiO2, 25 g) using 25% EtOAc in hexanes as the
eluent to afford 31 as a colorless oil (2.44 g, 85%, 95:1 d.r): Rf )
0.37 (1:1, Hex:EtOAc); 1H NMR (CDCl3) δ 6.56 (s, 2H), 5.07 (q,
J ) 7.1 Hz, 1H), 4.14-4.13 (m, 2H), 3.81 (s, 6H), 3.77 (s, 3H),
2.42-2.36 (m, 1H), 2.12 (s, 1H), 1.47 (d, J ) 7.1 Hz, 3H), 0.88 (t,
J ) 7.1 Hz, 6H); 13C NMR (CDCl3) δ 174.6, 153.1, 135.8, 105.2,
63.1, 60.8, 59.1, 56.1, 42.8, 28.6, 19.7, 18.0, 14.7; HRFAB [M +
Li] 358.1852 (calculated C18H25NO6Li: 358.1842).
1
(4:1, Hex:EtOAc); H NMR (CDCl3) δ 6.75 (s, 2H), 5.01 (d, J )
2.2 Hz, 1H), 3.88 (s, 6H), 3.84 (s, 3H), 3.45 (s, 3H), 2.68 (d, J )
2.2 Hz, 1H); 13C NMR (CDCl3) δ 153.4, 138.2, 133.7, 104.4, 81.3,
76.0, 73.1, 61.0, 56.3, 56.2; HRFAB [M + Li] 243.1208 (calculated
C13H16O4Li: 243.1209).
2,4-Diamino-5-(3-methoxy-3-(3,4,5-trimethoxyphenyl)prop-1-
ynyl)pyrimidine (27). 5 (236 mg) was allowed to react with 26
(473 mg) as per the general procedure to afford 27 as an orange
powder (310 mg, 90%): Rf ) 0.31 (9:1, CHCl3:MeOH); mp )
1
184-186 °C; H NMR (DMSO-d6) δ 7.91 (s, 1H), 6.82 (s, 2H),
6.40 (s, 2H), 5.30 (s, 1H), 3.79 (s, 6H), 3.66 (s, 3H), 3.35 (s, 3H);
13C NMR (DMSO-d6) δ 163.8, 162.3, 159.7, 152.8, 137.2, 134.8,
104.6, 92.6, 89.1, 81.9, 73.0, 60.0, 55.9, 55.4; Anal. (C17H20N4O4)
C, H, N.
(S)-3-((S)-2-(3,4,5-Trimethoxyphenyl)propanoyl)-4-isopropy-
loxazolidin-2-o ne (32). 32 was prepared in an analogous manner
as 31. The residue was purified by flash chromatography (SiO2,
25 g) using 25% EtOAc in hexanes as the eluent to afford 32 as a
colorless oil (2.08 g, 87%, 95:1 d.r): Rf ) 0.37 (1:1, Hex:EtOAc);
1H NMR (CDCl3) δ 6.58 (s, 2H), 5.09 (q, J ) 7.1 Hz, 1H), 4.17-
4.15 (m, 2H), 3.83 (s, 6H), 3.80 (s, 3H), 2.45-2.39 (m, 1H), 2.15
(s, 1H), 1.49 (d, J ) 7.1 Hz, 3H), 0.90 (t, J ) 7.1 Hz, 6H); 13C
NMR (CDCl3) δ 174.7, 153.2, 135.9, 105.3, 63.2, 60.9, 59.2, 56.2,
42.9, 28.6, 19.8, 18.1, 14.8; HRFAB [M + Li] 358.1856 (calculated
C18H25NO6Li: 358.1842).
General Procedure for the Synthesis of 35 and 36. To a flame-
dried 200 mL round-bottom flask was added the desired oxazoli-
dinone (1.0 equiv). DCM (0.1 M) was added, and the solution was
cooled to -78 °C. DIBAL-H (2.0 equiv) was added dropwise, and
the reaction was allowed to stir at -78 °C for 2 h. Saturated NH4-
Cl (20 mL) was added and the solution warmed to room temper-
ature. The solids were filtered off and washed with DCM (3 × 5
mL). The aqueous layer was extracted with EtOAc (3 × 10 mL).
The combined organics were washed with brine (10 mL), dried
over anhydrous Na2SO4, and concentrated. An NMR measurement
of the crude material was taken, and the spectrum matched that for
the racemic aldehyde 18. The crude material was used without
further purification.
CBr4 (1.5 equiv) was dissolved in DCM (0.1 M) and cooled to
0 °C. Ph3P (3.0 equiv) was added and the solution stirred at 0 °C
for 5 min. The crude aldehyde (1.0 equiv) in DCM (5.0 mL) was
added dropwise. The reaction was allowed to stir at 0 °C for 30
min and then poured into ice-cooled Et2O (200 mL). The reaction
was filtered through silica gel and washed with hexanes (100 mL)
followed by 25% EtOAc in hexanes (300 mL). The organics were
combined and concentrated under reduced pressure. The crude
material was taken on to the next step without further purification.
Mg (2.0 equiv) was suspended in THF (1.0 mL). Dibromoethane
(0.4 equiv) was added, and the suspension was stirred at 25 °C for
30 min. The crude dibromide (1.0 equiv) in THF (6.0 mL) was
added dropwise, and the reaction was heated at reflux for 30 min.
The solution was cooled to room temperature and the solvent
removed. The residue was purified by flash chromatography to
afford the respective enantioenriched acetylenes. NMR spectra were
taken, and the spectra matched that for the racemic acetylene 19.
2,4-Diamino-5-((R)-3-(3,4,5-trimethoxyphenyl)but-1-ynyl)-6-
methylpyrimidine (37). 31 (2.44 g, 6.94 mmol) was subjected to
the general procedure to afford 35 (460 mg, 30%). 15 (250 mg)
was allowed to react with 35 (290 mg) as per the general
Sonagashira coupling procedure to afford 37 as a brown powder
(280 mg, 82%, 90% ee): Rf ) 0.29 (9:1, CHCl3:MeOH); 1H NMR
(DMSO-d6) δ 6.76 (s, 2H), 6.19 (s, 2H), 4.02 (q, J ) 7.1 Hz, 1H),
3.77 (s, 6H), 3.63 (s, 3H), 2.21 (s, 3H), 1.51 (d, J ) 7.1 Hz, 3H);
13C NMR (DMSO-d6) δ 166.9, 164.1, 160.9, 152.8, 139.4, 135.9,
104.0, 100.7, 88.6, 76.4, 60.0, 55.8, 32.5, 24.7, 22.5; HRFAB [M
+ Li] 349.1851 (calculated C18H22N4O3Li: 349.1851).
2,4-Diamino-5-(3-methoxy-3-(3,4,5-trimethoxyphenyl)prop-1-
ynyl)-6-methylpyrimidine (28). 15 (250 mg) was allowed to react
with 26 (473 mg) as per the general procedure to afford 28 as a
yellow powder (335 mg, 93%): Rf ) 0.29 (9:1, CHCl3:MeOH);
1
mp ) 127-129 °C; H NMR (DMSO-d6) δ 6.84 (s, 2H), 6.37 (s,
2H), 5.34 (s, 1H), 3.78 (s, 6H), 3.66 (s, 3H), 3.35 (s, 3H), 2.23 (s,
3H); 13C NMR (DMSO-d6) δ 167.6, 164.3, 161.1, 152.8, 137.2,
134.8, 104.6, 95.7, 87.5, 81.8, 73.1, 60.0, 55.8, 55.3, 22.4; HRFAB
[M + Li] 359.1701 (calculated C18H22N4O4Li: 359.1719).
(R)-4-Isopropyl-3-(2-(3,4,5-trimethoxyphenyl)acetyl)oxazolidin-
2-one (29). To a flame-dried 100 mL round-bottom flask was added
3,4,5-trimethoxyphenylacetic acid (2.10 g, 9.29 mmol). THF (25
mL) was added followed by Et3N (1.42 mL, 10.22 mmol). The
solution was cooled to -78 °C. Pivaloyl chloride (1.26 mL, 10.22
mmol) was added dropwise and the solution warmed to 0 °C and
stirred for 1 h. In a separate flame-dried 50 mL round-bottom flask
was added (R)-4- isopropyloxazolidin-2-one (1.0 g). THF (20 mL)
was added and the solution cooled to -78 °C. n-BuLi (6.83 mL,
9.29 mmol, 1.36 M) was added dropwise and the solution stirred
at -78 °C for 15 min and then warmed to 25 °C where it was
stirred for 15 min. The organolithium solution was transferred to
the solution of the mixed anhydride via cannula at -78 °C. The
reaction was stirred at -78 °C for 15 min, warmed to 0 °C, and
stirred for 1 h. Water (10 mL) was added and the aqueous layer
extracted with EtOAc (2 × 10 mL). The combined organics were
washed with brine (10 mL), dried over anhydrous MgSO4, and
concentrated under reduced pressure. The residue was purified by
flash chromatography (SiO2, 30 g) using 25% EtOAc in hexanes
as the eluent to afford 29 as a colorless oil (2.30 g, 88%): Rf )
1
0.28 (1:1, Hex:EtOAc); H NMR (CDCl3) δ 6.55 (s, 2H), 4.44-
4.41 (m, 1H), 4.29-4.24 (m, 2H), 4.19 (dd, J ) 9.0, 3.0 Hz, 1H),
4.12-4.07 (m, 1H), 3.82 (s, 6H), 3.80 (s, 3H), 2.36-2.30 (m, 1H),
0.87 (d, J ) 6.8 Hz, 3H), 0.78 (d, J ) 7.1 Hz, 3H); 13C NMR
(CDCl3) δ 171.2, 154.1, 153.2, 129.4, 106.7, 63.4, 60.9, 58.6, 56.2,
41.6, 28.4, 18.0, 14.7, 14.3; HRFAB [M + Li] 344.1686 (calculated
C17H23NO6Li: 344.1686).
(S)-4-Isopropyl-3-(2-(3,4,5-trimethoxyphenyl)acetyl)oxazolidin-
2-one (30). 30 was synthesized in an analogous manner as 29 using
(S)-4-isopropyloxazolidin-2-one. The residue was purified by flash
chromatography (SiO2, 30 g) using 25% EtOAc in hexanes as the
eluent to afford 30 as a colorless oil (2.30 g, 88%): Rf ) 0.28
1
(1:1, Hex:EtOAc); H NMR (CDCl3) δ 6.54 (s, 2H), 4.44-4.41
(m, 1H), 4.29-4.24 (m, 2H), 4.19 (dd, J ) 9.0, 3.0 Hz, 1H), 4.12-
4.07 (m, 1H), 3.82 (s, 6H), 3.80 (s, 3H), 2.36-2.30 (m, 1H), 0.87
(d, J ) 6.8 Hz, 3H), 0.78 (d, J ) 7.1 Hz, 3H); 13C NMR (CDCl3)
δ 171.2, 154.1, 153.2, 129.4, 106.7, 63.4, 60.9, 58.6, 56.2, 41.6,
28.4, 18.0, 14.6, 14.3; HRFAB [M + Li] 344.1700 (calculated
C17H23NO6Li: 344.1686).
(R)-3-((R)-2-(3,4,5-trimethoxyphenyl)propanoyl)-4-isopropy-
loxazolidin-2-o ne (31). To a flame-dried 200 mL round-bottom
flask was added 29 (2.77 g, 8.21 mmol). THF (85 mL) was added
and the solution cooled to -78 °C. LHMDS (12.5 mL, 12.32 mmol,
1.0 M) was added dropwise, and the reaction was allowed to stir
at -78 °C for 1 h. MeI (1.54 mL, 24.63 mmol) was added and the
solution stirred at -78 °C for 1 h. The solution was then warmed
to 0 °C for 1 h and quenched with sat. NH4Cl (10 mL). The aqueous
2,4-Diamino-5-((S)-3-(3,4,5-trimethoxyphenyl)but-1-ynyl)-6-
methylpyrimidine (38). 32 (1.66 g, 4.72 mmol) was subjected to
the general procedure to afford 36 (312 mg, 30%). 15 (250 mg)
was allowed to react with 36 (290 mg) as per the general
Sonagashira coupling procedure to afford 38 as a yellow powder
(308 mg, 90%, 95% ee): Rf ) 0.29 (9:1, CHCl3:MeOH); 1H NMR
(DMSO-d6) δ 6.76 (s, 2H), 6.19 (s, 2H), 4.02 (q, J ) 7.1 Hz, 1H),
3.77 (s, 6H), 3.63 (s, 3H), 2.21 (s, 3H), 1.51 (d, J ) 7.1 Hz, 3H);

950 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 5
Pelphrey et al.
13C NMR (DMSO-d6) δ 167.0, 164.1, 161.0, 152.8, 139.4, 136.0,
104.0, 100.8, 88.6, 76.4, 60.0, 55.8, 32.5, 24.6, 22.5; HRFAB [M
+ Li] 349.1851 (calculated C18H22N4O3Li: 349.1851).
(16) Champness, J.; Achari, A.; Ballantine, S.; Bryant, P.; Delves, C.;
Stammers, D. The structure of Pneumocystis carinii Dihydrofolate
Reductase to 1.9 Å Resolution. Structure 1994, 2, 915-924.
(17) Knighton, D.; Kan, C.; Howland, E.; Janson, C.; Hostomska, Z.;
Welsh, K.; Matthews, D. Structure of and kinetic channeling in
bifunctional dihydrofolate reductase-thymidylate synthase. Nat. Struct.
Biol. 1994, 1, 186-194.
(18) Jones, M. L.; Baccanari, D. P.; Tansik, R. L.; Boytos, C. M.; Rudolph,
S. K.; Kuyper, L. F. Inhibitors of dihydrofolate reductase: Design,
synthesis and antimicrobial activities of 2,4-diamino-6-methyl-5-
ethynylpyrimidines. J. Heterocycl. Chem. 1999, 36 (1), 145-148.
(19) deSousa, P. T. The palladium-catalysed cross coupling reaction of
acetylenic compounds with aryl halides and related compounds.
Quim. NoVa 1996, 19 (4), 377-382.
(20) Muller, S.; Liepold, B.; Roth, G. J.; Bestmann, H. J. An improved
one-pot procedure for the synthesis of alkynes from aldehydes. Synlett
1996, 6, 521-522.
(21) McMartin, C.; Bohacek, R. QXP: Powerful, rapid computer algo-
rithms for structure-based drug design. J. Comput.-Aided Mol. Des.
1997, 11, 333-344.
(22) Jain, A., Surflex: Fully Automatic Flexible Molecular Docking Using
a Molecular Similarity-Based Search Engine. J. Med. Chem. 2003,
46, 499-511.
(23) Jain, A. Ligand-Based Structural Hypotheses for Virtual Screening.
J. Med. Chem. 2004, 47, 947-961.
(24) Ruppert, J.; Welch, W.; Jain, A. Automatic identification and
representation of protein binding sites for molecular docking. Protein
Sci. 1997, 6, 524-533.
(25) Welch, W.; Ruppert, J.; Jain, A. Hammerhead: fast, fully automated
docking of flexible ligands to protein binding sites. Chem. Biol. 1996,
3, 449-462.
(26) Jain, A. Scoring noncovalent protein-ligand interactions: a continuous
differentiable function tuned to compute binding affinities. J.
Comput.-Aided Mol. Des. 1996, 10, 427-440.
(27) Danishefsky, S.; Harvey, D. A new approach to polypropionates:
routes to subunits of monensin and tirandamycin. J. Am. Chem. Soc.
1985, 107, 6647-6652.
(28) Van Hijfte, L.; Kolb, M.; Witz, P. A practical procedure for the
conversion of aldehydes to terminal alkynes by a one carbon
homologation. Tetrahedron Lett. 1989, 30, 3655-3656.
(29) Bull, S.; Davies, S.; Key, M.; Nicholson, R.; Savory, E. Conforma-
tional control in the SuperQuat chiral auxiliary 5,5-dimethyl-4-iso-
propyloxazolidin-2-one induces the iso-propyl group to mimic a tert-
butyl group. Chem. Commun. 2000, 18, 1721-1722.
(30) Rosowsky, A.; Forsch, R.; Queener, S. Further Studies on 2,4-
Diamino-5-(2′,5′-disubstituted benzyl)pyrimidines as Potent and
Selective Inhibitors of Dihydrofolate, Reductases from Three Major
Opportunistic Pathogens of AIDS. J. Med. Chem. 2003, 46, 1726-
1736.
Acknowledgment. The authors acknowledge Dr. Robin
Couch for assistance with chiral chromatography and support
of the NIH (GM067542 to A.C.A.) and NSF (#013468 to
A.C.A.) for this work.
Supporting Information Available: Details of the preparation
of TgDHFR as well as spectra for each compound. This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Anderson, A. Targeting DHFR in parasitic protozoa. Drug DiscoVery
Today 2005, 10, 121-128.
(2) Chio, L.-C.; Queener, S. Identification of Highly Potent and Selective
Inhibitors of Toxoplasma gondii Dihydrofolate Reductase. Antimi-
crob. Agents Chemother. 1993, 37, 1914-1923.
(3) 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
structures. Nat. Struct. Biol. 2003, 10, 357-365.
(4) Popov, V.; Chan, D.; Fillingham, Y.; Yee, W. A.; Wright, D.;
Anderson, A. Analysis of complexes of inhibitors with Cryptospo-
ridium hominis DHFR leads to a new trimethoprim derivative.
Bioorg. Med. Chem. Lett. 2006, 16, 4366-4370.
(5) Rosowsky, A.; Forsch, R.; Queener, S. Inhibition of Pneumocystis
carinii Toxoplasma gondii and Mycobacterium avium Dihydrofolate
Reductases by 2,4-Diamino-5-[2-methoxy-5-(ω-carboxyalkoxy)ben-
zyl]pyramidines: Marked Improvement in Potency Relative to
Trimethoprim and Species Selectivity Relative to Piritrexim. J. Med.
Chem. 2002, 45, 233-241.
(6) Trujillo, M.; Donald, R.; Roos, D.; Greene, P.; Santi, D. Heterologous
Expression and Characterization of Bifunctional Dihydrofolate Re-
ductase-Thymidylate Synthase Enzyme of Toxoplasma gondii.
Biochemistry 1996, 35, 6366-6374.
(7) Lipinski, C.; Lombardo, F.; Dominy, B.; Feeney, P. Experimental
and computational approaches to estimate solubility and permeability
in drug discovery and development settings. AdV. Drug DeliVery ReV.
1997, 23, 3-25.
(8) Wenlock, M.; Austin, R.; Barton, P.; Davis, A.; Leeson, P. A
comparison of physiochemical property profiles of development and
marketed oral drugs. J. Med. Chem. 2003, 46, 1250-1256.
(9) Hopkins, A.; Groom, C.; Alex, A. Ligand efficiency: a useful metric
for lead selection. Drug DiscoVery Today 2004, 9, 430-431.
(10) Kuntz, I.; Chen, K.; Sharp, K.; Kollman, P. The maximal affinity of
ligands. Proc. Natl. Acad. Sci. 1999, 96, 9997-10002.
(11) Popov, V.; Yee, W. A.; Anderson, A. Towards in silico lead
optimization: scores from ensembles of protein:ligand conformations
reliably correlate with biological activity Proteins 2006, 66, 375-
387.
(31) Rosowsky, A.; Fu, H.; Chan, D.; Queener, S., Synthesis of 2,4-
Diamino-6-[2′-O-(ω-carboxyalkyl)oxydibenz[b,f]azepin-5-yl]meth-
ylpt eridines as Potent and Selective Inhibitors of Pneumocystis
carinii, Toxoplasma gondii, and Mycobacterium aVium Dihydrofolate
Reductase. J. Med. Chem. 2004, 47, 2475-2485.
(32) Gangjee, A.; Devraj, R.; Queener, S. Synthesis and dihydrofolate
reductase inhibitory activity of 2,4-diamino-5-deaza and 2,4-diamino-
5,10-dideaza lipophilic antifolates. J. Med. Chem. 1997, 40, 470-
478.
(33) Gangjee, A.; Adair, O.; Queener, S., Synthesis and biological
evaluation of 2,4-diamino-6-(arylaminomethyl)pyrido[2,3-d]pyrim-
idines as inhibitors of Pneumocystis carinii and Toxoplasma gondii
dihydrofolate reductase and as antiopportunistic infection and
antitumor agents. J. Med. Chem. 2003, 46, 5074-5082.
(34) Rosowsky, A.; Chen, H.; Fu, H.; Queener, S. Synthesis of new 2,4-
diaminopyrido[2,3-d]pyrimidine and 2,4-diaminopyrrolo[2,3-d]py-
rimidine inhibitors of Pneumocystis carinii, Toxoplasma gondii and
Mycobacterium avium dihydrofolate reductase. Bioorg. Med. Chem.
2003, 11, 59-67.
(12) 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.
(13) O’Neil, R.; Lilien, R.; Donald, B.; Stroud, R.; Anderson, A.
Phylogenetic classification of protozoa based on the structure of the
linker domain in the bifunctional enzyme, dihydrofolate reductase-
thymidylate synthase. J. Biol. Chem. 2003, 278, 52980-52987.
(14) Cody, V.; Luft, J.; Pangborn, W. Understanding the role of Leu22
variants in methotrexate resistance: comparison of wild-type and
Leu22Arg, variant mouse and human dihydrofolate reductase ternary
crystal complexes with methotrexate and NADPH. Acta Crystallogr.
2005, D61, 147-155.
(15) 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.
JM061027H