Discovery of Potent Inhibitors of Dihydroneopterin Aldolase Using CrystaLEAD High-Throughput X-ray Crystallographic Screening and Structure-Directed Lead Optimization

Article abstract of DOI:10.1021/jm030497y

Potent inhibitors of 7,8-dihydroneopterin aldolase (DHNA; EC 4.1.2.25) have been discovered using CrystaLEAD X-ray crystallographic high-throughput screening followed by structure-directed optimization. Screening of a 10 000 compound random library provid

Full text of DOI:10.1021/jm030497y

J . Med. Chem. 2004, 47, 1709-1718
1709
Discover y of P oten t In h ibitor s of Dih yd r on eop ter in Ald ola se Usin g Cr ysta LEAD
High -Th r ou gh p u t X-r a y Cr ysta llogr a p h ic Scr een in g a n d Str u ctu r e-Dir ected
Lea d Op tim iza tion
William J . Sanders,* Vicki L. Nienaber, Claude G. Lerner, J . Owen McCall, Sean M. Merrick,
Susan J . Swanson, J ohn E. Harlan, Vincent S. Stoll, Geoffrey F. Stamper, Stephen F. Betz, Kevin R. Condroski,
Robert P. Meadows, J ean M. Severin, Karl A. Walter, Peter Magdalinos, Clarissa G. J akob, Rolf Wagner, and
Bruce A. Beutel
Infectious Disease Research, Abbott Laboratories, 200 Abbott Park Road, Abbott Park, Illinois 60064-6217
Received October 3, 2003
Potent inhibitors of 7,8-dihydroneopterin aldolase (DHNA; EC 4.1.2.25) have been discovered
using CrystaLEAD X-ray crystallographic high-throughput screening followed by structure-
directed optimization. Screening of a 10 000 compound random library provided several low
affinity leads and their corresponding X-ray crystal structures bound to the enzyme. The
presence of a common structural feature in each of the leads suggested a strategy for the
construction of a directed library of approximately 1000 compounds that were screened for
inhibitory activity in a traditional enzyme assay. Several lead compounds with IC50 values of
about 1 µM against DHNA were identified, and crystal structures of their enzyme-bound
complexes were obtained by cocrystallization. Structure-directed optimization of one of the leads
thus identified afforded potent inhibitors with submicromolar IC50 values.
In tr od u ction
catalyzes the retroaldol reaction of 7,8-dihydroneopterin
(3) to afford glycoaldehyde and 6-hydroxymethyl-7,8-
dihydropterin (4) (Figure 1).4,7 Like DHPS and DHFR,
DHNA is conserved across bacterial species and is
nonexistent in humans. Unlike the former, however,
DHNA has not been extensively investigated as a
potential therapeutic target.8
The increasing prevalence of microorganisms display-
ing resistance to known antibiotics has fueled continued
interest in the discovery of potent, broad-spectrum
antibacterial agents.1 Among the most promising strat-
egies for the development of new antibacterial thera-
peutics is the targeting of proteins essential for bacterial
growth but lacking mammalian counterparts.2 Charac-
teristics of an ideal therapeutic target protein include
broad conservation in the structure, or at least in the
active site structure, across relevant bacterial strains.
In addition, the lack of eukaryotic orthologues provides
a greatly reduced chance of side effects induced by
inhibition of similar proteins in humans.
The folate biosynthetic pathway has long been rec-
ognized as an attractive target for the development of
selective antimicrobial agents.3 Folate derivatives are
essential cofactors in the synthesis of purines, pyri-
midines, and amino acids in most organisms. Unlike
humans and other mammals, bacteria are unable to
absorb folate from their environment and must there-
fore synthesize it de novo. The biosynthetic pathway for
the production of tetrahydrofolate (Figure 1) requires
enzymes essential for bacterial growth, several of which
do not possess eukaryotic orthologues.4 A number of
successful, marketed antibacterials including the sul-
fonamides (dihydropteroate synthase, DHPS)5 and tri-
methoprim (dihydrofolate reductase, DHFR)6 target
enzymes in this pathway. DHPS and DHFR both
possess the desired attributes of a successful antibacter-
ial target: broad species homology and a lack of closely
related enzymes in mammals. A third enzyme in the
pathway, dihydroneopterin aldolase (DHNA; EC 4.1.2.25),
X-r a y Cr ysta llogr a p h y
The crystal structure of DHNA from Staphylococcus
aureus, first reported by Hennig and co-workers,9 shows
that this enzyme crystallizes as an octamer of I422
symmetry. Specifically, the octamer is composed of two
donut-shaped tetramers that come together to form a
cylinder. The active site is located at the interface
between monomer subunits of the tetramer such that
one part of the active site is formed by one monomer
while the other is formed by a second monomer, result-
ing in four active sites per tetramer (Figure 2). The
crystal structure of DHNA in complex with the product
of the enzyme-catalyzed reaction, 6-hydroxymethyl-7,8-
dihydropterin 4 (Figure 3), was also reported by Hennig9
and confirmed the location of the enzyme active site.
Many qualities of the reported DHNA crystal form
make it a good candidate for lead discovery using high-
throughput X-ray crystallographic screening.10 The
DHNA space group is high symmetry, thus requiring,
on average, fewer data frames for a complete sampling
of the unique section of the X-ray diffraction reciprocal
lattice. This fact, coupled with the reported high-
resolution diffraction of the crystals, suggests that data
collection times should be relatively short on a high-
intensity rotating anode source equipped with a detector
with a short read-out time such as a charged-coupled
device (CCD) detector.11 Furthermore, because it was
demonstrated that the ligand could be diffused into the
* To whom correspondence should be addressed. Phone: (847) 937-
7168. Fax: (847) 935-5165. E-mail: will.sanders@abbott.com.
10.1021/jm030497y CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/18/2004

1710 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7
Sanders et al.
F igu r e 1. De novo biosynthesis of tetrahydrofolate. Enzymatic targets of marketed antibacterials sulfamethoxazole (9) and
trimethoprim (10) are indicated with arrows.
F igu r e 3. Structures of the DHNA native substrate 7,8-
dihydroneopterin (3), substrate analogue neopterin (11), and
the product of the enzyme-catalyzed reaction (4).
F igu r e 2. Ribbon diagram of dihydroneopterin aldolase
showing the tetrameric structure with the four catalytic sites
at monomer interfaces highlighted.
cannot occur in the crystalline state. An inability to
make conformational transitions may cause crystal
cracking upon treatment with ligand, resulting in
crystals that do not diffract. Alternatively, diffraction-
quality crystals with no ligand bound may result,
leading to false negative binding data. However, because
of our requirement for a lead compound and because
published reports indicated that the natural substrate
could be bound and that turnover could occur within
the crystal lattice, crystallographic screening was per-
formed with the possibility of false negative results
taken into consideration.
active site of intact DHNA crystals,9 crystallographic
screening for lead compound fragments (CrystaLEAD)
was an ideal choice to screen for inhibitors of DHNA.10
The combination of straightforward crystal production,
a short acquisition time for diffraction data, and the
ability to diffuse ligands into the active site of the intact
crystal make DHNA a very attractive target for Crys-
taLEAD screening. A potential drawback to the use of
crystallographic screening, however, is the location of
the DHNA active site. Because the enzyme active sites
are located at subunit interfaces within a tetramer, it
is possible that the protein may undergo allosteric
conformational transitions upon ligand binding that
The ability to diffuse ligands into an enzyme active
site in its crystalline state is the preferred method for
finding leads using the crystallographic screening method
CrystaLEAD developed by Nienaber and co-workers.10

Inhibitors of Dihydroneopterin Aldolase
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7 1711
F igu r e 4. Model of neopterin bound in the active site of
DHNA with key hydrogen bonds shown as dashed lines. The
top right of the picture depicts an unoccupied groove that could
be accessed to increase the potency of inhibitors.
Hence, although it was reported by Hennig and co-
workers9 that diffusion of ligands into DHNA crystals
was successful, the uncleavable substrate analogue
neopterin 11 (Figure 3) and the natural substrate 7,8-
dihydroneopterin 3 (data not shown) were soaked into
the crystals as positive controls. In both experiments,
the diffused compound was clearly visible in the initial
electron density maps (2Fo - Fc contoured at 1σ and Fo
- Fc contoured at 2.5σ). As reported by Hennig, the
bicyclic aromatic core of neopterin is highly ordered,
while the terminus of the aliphatic triol substituent
becomes disordered in the electron density map. Ex-
amination of the neopterin/DHNA complex structure
shows a binding mode for the inhibitor that is similar
to the product complex previously reported.9 Notably,
there appear to be three polar groups on the inhibitor
that are engaged in hydrogen bonds with the protein
at the monomer-monomer interface (Figure 4). This
gives rise to a three-point hydrogen-bonding motif
between the inhibitor and the protein that is probably
critical for substrate recognition. Specifically, hydrogen
bonds exist between the 4-carbonyl of the ligand and
the backbone nitrogen of Leu 73 of one of the monomers
(molecule A) as well as between the 3-position -NH of
the inhibitor and Glu 74-Oδ1 of molecule A. The
exocyclic 2-amino group of the ligand is donating
hydrogen bonds to Glu 74-Oδ2 of molecule A and to Val
52-O of the second protein monomer (molecule B, not
shown). The ordered solvent molecule reported in the
literature structure9 that is hydrogen-bonding with Asn
71-O, Lys 100-Nú (both in molecule A), and the 4-car-
bonyl of the inhibitor is also observed in the neopterin
complex structure. Interestingly, the substrate binding
pocket shows some similarity to the primary binding
pocket of trypsin-like proteases.12 Glu 74, which is
located at the base of the DHNA pocket, is analogous
to Asp 189 of the trypsin-like proteases. This similarity
suggests that screening of directed libraries designed
to identify inhibitors of trypsin-like enzymes could be a
source of lead compounds for DHNA as well. Upon
completion of exploratory experiments and evaluation
of the DHNA active site structure, crystallographic
screening for lead compounds was initiated.
F igu r e 5. Results of crystallographic screening experiments.
All maps shown are the initial 2Fo - Fc maps calculated for
the protein in the absence of inhibitor. These are the maps
that were examined for analysis of the crystallographic screen-
ing data: (A) 9-methylguanine (12, IC50 ) 28 µM) at 2.2 Å
resolution; (B) 2-amino-4-hydroxy-5-carboxyethylpyrimidine
(13, IC50 ) 80 µM) at 2.2 Å resolution; (C) 8-amino-1,3-
dimethyl-3,7-dihydropurine-2,6-dione (14, IC50 ) 50 µM) at 2.3
Å resolution; (D) 2-amino-5-bromo-3-hydroxy-6-phenylpyrimi-
dine (15, IC50 ) 38 µM) at 2.1 Å resolution. Hydrogen bonds
between each inhibitor and DHNA are depicted by dashed
lines.
density map of a bound complex. The sublibraries were
then screened as mixtures, resulting in the identifica-
tion of two ligands bound to DHNA (parts A and B of
Figure 5). Both compounds occurred in the same 100-
compound sublibrary and were unambiguously identi-
fied from the mixture without deconvolution based on
the shape of the initial electron density maps.10 Both
hits shared the three-point hydrogen-bonding motif of
the substrate analogue depicted in Figure 4 and were
subsequently shown to be competitive inhibitors of
DHNA. 9-Methylguanine 12 (Figure 5A) was found to
inhibit DHNA in the enzymatic assay with an IC50 of
28 µM. Compound 13, a 2-aminopyrimidine (Figure 5B),
similarly inhibited the enzyme but was less potent with
an IC50 of 80 µM. Because of the similarity between the
DHNA binding pocket and the primary binding site (S1)
of trypsin-like serine proteases mentioned above, a
fragment library designed to be screened against the
trypsin-like protease urokinase10 was also screened for
DHNA binders. This screen resulted in identification
of compound 14, an 8-aminopurine analogue (Figure
5C). Because the 8-aminopurine analogue is symmetric,
the molecule can be fitted into the electron density in
two possible orientations. The orientation shown in
Figure 5C is the conformation in which the inhibitor
maintains a hydrogen bond with Leu 73-N as in the
substrate/enzyme complex. Compound 14 was also
A 10 000 compound library was divided into 100-
compound sublibraries with sufficient structural diver-
sity to allow for direct hit identification from the electron

1712 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7
Sanders et al.
F igu r e 7. Representative enzyme inhibition curves. The rate
of the enzymatic reaction is plotted vs increasing inhibitor
concentration, and IC50 is determined by the concentration of
compound that causes a 50% decrease in the initial reaction
rate. The curve observed for 11 is typical of weak inhibitors,
while the curve observed for 19 shows a much steeper slope,
typical of more potent inhibitors.
indicating a possible conformational shift during ligand
binding. Examples of the two types of inhibition curves
observed for DHNA inhibitors are illustrated in Figure
7. The moderately sloped curve of weak inhibitors is
represented by neopterin 11, while the steeply sloped
curve of carboxylic acid 19 is representative of more
potent inhibitors. Since a conformational shift during
ligand binding was suspected, individual cocrystalliza-
tion experiments were performed in the presence of a
number of the lead compounds identified in the directed
library screen. Almost all of the compounds were suc-
cessfully cocrystallized, yielding isomorphous crystals
(space group I422) similar to the crystals obtained by
the crystal soak method. One compound, however,
afforded crystals in which a space group shift to P4(2)-
22 with a dimer rather than a monomer in the asym-
metric unit occurred. Still, close visual comparison of
each monomer of this dimer to the I422 monomer
showed no significant difference in the crystal lattices.
These results suggest that a conformational shift that
cannot occur in the crystal lattice does transpire during
compound binding; however, the final bound structure
remains quite similar to the apo structure. These data
also suggest that binding outside the primary binding
site and into an extended groove illustrated in Figure 4
could trigger this shift. Hence, it is possible that
compounds present in the original 10 000 compound
library could require a similar conformational shift and
would therefore not bind to the crystalline state of
DHNA. Although the possibility exists that potential
leads were overlooked because of the inability of the
crystal soaking method to identify compounds in which
a conformational transition occurs upon binding, the
variety and potency of those hits identified were suf-
ficient to provide promising leads for structure-based
optimization.
F igu r e 6. (A) Three-point hydrogen-bonding motif used in
the construction of a directed library for enzyme inhibition
screening. (B) Inhibitors identified in an enzyme inhibition
assay of a directed library.
shown to be a competitive inhibitor of DHNA with an
IC50 of 50 µM. Finally, a third library consisting solely
of aminopyrimidines was screened, resulting in a fourth
hit. Compound 15 (Figure 5D) was identified from its
X-ray diffraction pattern and was shown to competi-
tively inhibit DHNA with an IC50 of 38 µM.
Of the four lead scaffolds identified from the crystal-
lographic screens, three (12, 13, and 15) contained the
exact three-point hydrogen-bonding motif present in the
substrate analogue neopterin and in the bound product
6-hydroxymethyl-7,8-dihydropterin, and the fourth (14)
displayed hydrogen bonds to the same protein residues
but with different atom connectivity in the ligand. As
mentioned above, it was recognized early that false
negative binding data may result if an allosteric con-
formational shift occurred in the enzyme active site
upon ligand binding. The presence of an identical
hydrogen-bonding motif in three of the four CrystaL-
EAD hits and in both the substrate analogue and the
enzyme-catalyzed reaction product suggested that screen-
ing of a much smaller, directed library in an enzymatic
assay might prove to be fruitful. To this end, a library
consisting of ∼1000 compounds all containing the H2N-
C-NH-CdO substructure (Figure 6A) was constructed.
Screening of this directed library in the enzymatic assay
provided several hits with IC50 values on the order of 1
µM (Figure 6B) as well as a larger number with IC50
values from 10 to 50 µM (structures not shown).
Although the leads identified in the secondary screen
proved to be more potent than the original compounds
identified by CrystaLEAD, it was not possible to obtain
crystal structures by the compound soaking method.
Interestingly, all of these compounds displayed sigmoi-
dal inhibition curves in the enzyme assay (Figure 7),
consistent with cooperativity in the enzymatic assay and
Design , Syn th esis, a n d Activity of In h ibitor s
In the crystal structure of the neopterin/DHNA
complex, the neopterin heterocylic ring occupies most
of the volume of the DHNA substrate binding pocket.
Since this orientation proved to be identical for all of

Inhibitors of Dihydroneopterin Aldolase
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7 1713
Sch em e 1a
a
Reagents and conditions: (i) 1.4 M NaOAc, 91%; (ii) CuSO4,
pyridine, H2O, reflux, 85%; (iii) RR′NH, EDC, DIPEA, DMF.
Ta ble 1. Side Chain Structures and Inhibitory Potencies of
Benzylic Amide Derivatives of Lead Compound 19
F igu r e 8. Molecular model of lead compound 19 bound to
DHNA constructed from the X-ray crystal structure of the
bound complex showing potential for access to an extended
binding groove. The protein surface is shown in blue with the
amino acids lining the groove (104-106 molecule A; 53, 55,
and 59 molecule B) highlighted in green and red. The binding
mode of the inhibitor is shown in yellow.
the lead molecules identified, additional nearby sites
were sought for exploitation in structure-based design
cycles to engineer both potency and specificity into
inhibitors. An examination of the active site and the
surrounding area identified an extended binding groove
above the heterocyclic ring, which could theoretically
be occupied by substitution of a ligand bound in the
primary binding pocket (Figure 4). This site is lined by
Pro 104, Ile 105, and Pro 106 of one monomer subunit
and by His 53, Gly 55, and Glu 59 of a second monomer
subunit and is partially occupied by the aliphatic triol
of neopterin in the substrate analogue complex. Hence,
a potential design strategy would be to access this
extended binding groove by chemical extension of a lead
bound in the primary pocket. The ideal lead compound
for DHNA would have a binding orientation that would
facilitate access to this extended binding site through
additional chemical elaboration. Of the four most potent
compounds discovered in the secondary enzymatic as-
say, carboxylic acid 19 appeared to be the most promis-
ing both in terms of its bound orientation and in its
obvious prearrangement for chemical modification. Fig-
ure 8 shows a surface area model of the DHNA active
site occupied by compound 19 (oxygen atoms of the
carboxylic acid moiety are not shown) constructed from
the X-ray structure of the bound complex. From the
structure, it is clear that the carbon atom of the
carboxylic acid group is pointing directly into the
unoccupied groove described above. The trajectory of the
carboxylic acid carbon is such that substitution with
functionality designed to fill the unoccupied groove
should afford new contacts with the enzyme and thus
increased potency of inhibitors.
available amines, amides of 19 were chosen as the initial
target compounds. Scheme 1 illustrates the synthetic
approach to these molecules. Carboxylic acid 19 was
synthesized as previously reported,13 with slight modi-
fications. 2,6-Diamino-4-hydroxypyrimidine 20 was
treated with 3-carboxybenzenediazonium tetrafluoro-
borate 2114 in aqueous sodium acetate to afford the
diazine 22 in 91% yield. Oxidative cyclization of 22 in
the presence of copper(II) sulfate afforded 19 in 85%
yield. Standard carbodiimide-promoted amide formation
then afforded the target compounds in good yield and
high purity.
Tables 1 and 2 depict a sampling of the compounds
synthesized by this method and their inhibitory activi-
ties against DHNA. The initial set of derivatives pre-
pared included a variety of benzylamides (Table 1) and
hydroxyalkylamides (Table 2) designed simply to test
the possibility of increasing inhibitor potency by filling
the open area surrounding the primary binding pocket.
A combination of hydrophobic and hydrophilic substit-
uents was included to maximize the chances for in-
Synthetic opportunities for the functionalization of a
carboxylic acid moiety abound. Through simple, well-
precedented chemistry, carboxylic acid 19 could be
easily converted into a library of amides, esters, amines,
or ethers as well as a wide variety of more complex
functional groups. In this case, since the initial objective
was to discover space-filling groups to enhance the
potency of inhibitors, the carboxylic acid was viewed
simply as a linker to which to append a variety of space-
filling substituents. Because of the simplicity of amide
bond formation and the wide variety of commercially

1714 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7
Sanders et al.
Ta ble 2. Side Chain Structures and Inhibitory Potencies of
Hydroxyalkyl Amide Derivatives of Lead Compound 19
F igu r e 9. Enzyme inhibition curves for lead compound 19
and for compound 27, the most potent inhibitor identified.
creased potency through new ligand-enzyme interac-
tions. Although the native substrate and the product of
the enzyme-catalyzed reaction both contain hydroxyl
groups, the nature of the extended binding groove is
primarily hydrophobic, suggesting that hydrophobic
substituents would provide the largest potency boost.
Additionally, both small and large substituents were
included because of evidence of conformational trans-
formation upon binding. The possibility of a change in
the position of active site residues such as His 53, which
was observed in the X-ray structures of several related
compounds (data not shown) resulting in significant
enlargement of the binding pocket, required the inclu-
sion of amide substituents that qualitatively appeared
too large to fit into the cavity.
As expected and as the data in Tables 1 and 2
illustrate, the potency of inhibitors is increased by filling
the open space in the extended groove protruding from
the active site. The most potent compound prepared in
the current study is 3,5-dichlorobenzyl amide 27 with
an IC50 of 68 nM. Figure 9 shows the inhibition curve
of 27 along with that of the lead compound 19. As
described above, the curve is sigmoidal in shape,
indicating a possible conformational transition upon
binding. The variety of sizes and shapes of side chain
substituents investigated here provides a limited picture
of the nature of the newly identified binding site. The
variation of IC50 values among similarly substituted
compounds suggests a strong SAR within the binding
site; however, a much larger set of analogues is required
to more thoroughly probe the nature of this SAR. Of
particular interest are compounds 25 and 27 (Table 1).
Compound 27, the most potent inhibitor of DHNA
F igu r e 10. Molecular models of compounds 27 and 29 bound
to DHNA constructed from the X-ray crystal structures show-
ing occupation of the extended binding groove by aromatic
rings.
reported here, displays an IC50 that is 20-fold more
potent than the initial lead. The substitution of two
trifluoromethyl moieties for two chlorine atoms affords
compound 25, which is 15 times less potent than 27.
The observation of a large difference in potency between
two compounds of such similar size and shape indicates
that further optimization is quite feasible. Another
compound with promising indications for further opti-
mization is biphenyl thioether 29. Although not as
potent as 27, compound 29 is 3-fold more potent than
the lead. More importantly, the X-ray structure of 29
illustrates additional potential for structure-directed
optimization. X-ray structures of 27 and 29 are shown
in Figure 10 and illustrate a key feature of the initial
SAR. The biphenyl thioether substituent was chosen
because it qualitatively appeared too large to fit into
the binding site. However, the IC50 was surprisingly
potent, the reason for which is illustrated in the crystal
structure. In contrast to compound 27 (Figure 10A), the

Inhibitors of Dihydroneopterin Aldolase
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7 1715
conformation of 29 (Figure 10B) has shifted slightly in
the active site. While the core of the molecule remains
fixed, a slight rotation about the C(O)N-C bond pushes
the first phenyl ring lower and allows space for the
second phenyl ring to fit. The ability of the binding site
to adapt to conformational variation among ligands
combined with evidence of a strong SAR provides a solid
foundation for the identification of even more potent
inhibitors of DHNA.
with X-ray structures obtainable only by cocrystalliza-
tion. This result clearly illustrates a potential problem
with crystallographic screening and structure-directed
drug design in general. While the crystalline state of
an enzyme may select primarily for favorable conforma-
tions among compounds within a screening library, the
soluble enzyme can have entirely different require-
ments. In the case of DHNA, a large, extended binding
site that is not accessible within the crystalline state of
the enzyme becomes accessible through conformational
flexibility in the soluble enzyme. We have shown that
a significant improvement in IC50 values could be
realized by filling this extended binding site through
structure-directed design and anticipate further im-
provement with repeated design cycles.
Although one of the goals of this work was to identify
potent in vitro inhibitors of DHNA, the ultimate objec-
tive is the development of new antibacterial chemother-
apies. While structure-directed lead optimization led to
the synthesis of numerous compounds with IC50 values
less than 1 µΜ, including one with an IC50 of 68 nM,
none of the compounds prepared here inhibited the
growth of bacteria including Staphylococcus aureus,
Streptococcus pneumoniae, Moraxella catarrhalis, Hae-
mophilus influenzae, Enterococcus faecium, Streptococ-
cus pyogenes, and Staphylococcus epidermidis. In addi-
tion, the compounds were not active against a hypersu-
sceptible Escherichia coli permeability mutant and Acr-
efflux pump mutants of H. influenzae and E. coli. The
compounds were also inactive in the presence of perme-
ability-enhancing agents including EDTA, nisin, poly-
myxin B nanopeptide, and the efflux pump inhibitor
reserpine. Finally, none of the compounds described
here demonstrated the ability to inhibit bacterial uptake
of radiolabeled p-aminobenzoic acid, a key indicator to
the inhibition of folate biosynthesis.15 The surprising
lack of even weak antibacterial activity in multiple
assays including those performed in the presence of cell-
permeablizing agents and efflux inhibitors could have
several explanations. The use of permeablizers would
seem to rule out poor cell penetration as an issue, so it
is possible that inactivation of the inhibitors by intra-
cellular enzymes could be occurring or that nonspecific
binding to intracellular proteins or the cell membrane
may reduce the available inhibitor concentration. An
alternative and much more intriguing possibility is that
the intracellular concentration of the DHNA substrate
7,8-dihydroneopterin could be quite high relative to the
Km of the enzyme. In this case, even compounds with
submicromolar inhibitory activity are not potent enough
to decrease the intracellular folate pools to growth-
inhibitory levels. (It has been shown that upon treat-
ment with trimethoprim, intracellular folate pools drop
to 25% of uninhibited levels within 4 h.16) In any case,
we can conclude that although the current compounds
are potent inhibitors of DHNA in vitro, they do not show
promise as antibacterial therapeutics.
Exp er im en ta l Section
En zym e P r ep a r a tion s. The double-stranded DNA encod-
ing DHNA was isolated by PCR amplification from Staphylo-
coccus aureus genomic DNA using oligodeoxynucleotide prim-
ers (5′-GGAATTCACATATGCAAGACACAATCTTTCTTAAAGG
and 5′-GCGACGTACTCGAGTTATTTATTCTCCCTCACTA-
TTTCGATAC) based on the DHNA gene sequence (preliminary
sequence data were obtained from The Institute for Genomic
Research website at http://www.tigr.org). The gene was cloned
into the NdeI-XhoI restriction sites of the expression vector
pET-30b (Novagen, Madison, WI) behind the T7lac promoter.
A TAA-stop codon was inserted between the last codon of the
protein and any vector-derived C-terminal tail so that only the
native protein sequence was translated. All vector sequences
were confirmed by bidirectional dye-terminator DNA sequenc-
ing. The resulting construct was transformed into Escherichia
coli BL21(DE3) for expression. Large-scale expressions (17 L
culture volume) were carried out in a New Brunswick Scientific
(Edison, NJ ) Micros fermentor. Cells were grown in SBG
medium [per liter: 60 g of Superbroth (Q-Biogene, Carlsbad,
CA), 10 mL of 1% DF-60 antifoam, 10 mL of 30% glucose, 1
mL of trace element solution]. The fermentations were carried
out at 30 °C, and the pH was maintained at 7.00 through
automatic addition of the appropriate pH titrants (4 N KOH/4
N H2SO4). Air was sparged at a constant rate of 2 vvm (vvm
) liters of air per liter of medium per minute), and the
dissolved oxygen concentration was maintained at or above
45% air saturation throughout the entire run via a cascaded
control loop that automatically adjusted the agitation rate as
needed. A 30% glucose solution was fed during a portion of
the ferment to maintain the glucose concentration between 0.1
and 2 g/L. When the cultures reached an optical density (OD)
of 3.7, protein expression was induced by addition of 1 mM
IPTG. Cells were harvested 4 h postinduction with a resulting
cell yield of ∼27 g wet wt/L. The protein was expressed well
and was ∼80% soluble.
DHNA was purified by a modification of the procedure of
Hennig et al.9 Briefly, 20 g of cells were resuspended in buffer
A (25 mM Tris, pH 8.0 at 4 °C, 50 mM NaCl, 5% glycerol, 2
mM EDTA, 1 mM PMSF, 1 mM NaN3) and disrupted using a
French press. Following centrifugation, the supernatant was
loaded onto a 5 cm × 31 cm DEAE-Sepharose FF column
(Amersham Pharmacia Biotech, Picataway, NJ ). The protein
was eluted with a linear gradient of 0-1 M NaCl in buffer
A. The fractions containing the DHNA were pooled and
dialyzed into buffer A. The dialyzed pool was then loaded on
a 2.6 cm × 38 cm Q-Sepharose FF column (Amerscham
Pharmamcia Biotech, Picataway, NJ ) and eluted with a linear
gradient of 0-1 M NaCl in buffer A. The fractions containing
the DHNA were pooled and concentrated and then applied to
a 2.6 cm × 60 cm Sephacryl S300 HR column (Amerscham
Pharmamcia Biotech, Piscataway, NJ ) equilibrated in 25 mM
Tris, 5% glycerol, 50 mM NaCl, and 1 mM NaN3 (buffer B).
The purified DHNA was pooled and filter-sterilized. Recovery
averaged about 6 mg of purified protein per gram of cell paste.
Con clu sion
We have reported an example of high-throughput
X-ray crystallographic screening for an apparently
allosteric enzyme. These data seem to show that a
conformational shift that is not achievable in the
crystalline state of the enzyme is required for the
binding of potent inhibitors to DHNA, although the
bound and unbound enzyme conformations are identical.
While X-ray crystallographic screening of random li-
braries afforded inhibitors with weak IC50 values in the
range 20-80 µM, a traditional enzyme screen of a
directed library afforded much more potent inhibitors

1716 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7
Sanders et al.
Material thus purified routinely crystallized, with the resulting
crystals defracting to better than 2.5 Å.
desorption chemical ionization (DCI) mass spectrometry were
performed on Finnigan SSQ700 single quadrupole mass
spectrometers. Elemental analyses were performed by Rob-
ertson Microlit Laboratories (Madison, NJ ). All of the com-
pounds reported below were analyzed by LC-MS, 1H NMR,
and mass spectrometry and are consistent with the proposed
structures in each case. Analytical purity was assessed by
reverse-phase HPLC on a Phenomenex Luna C8 column (5 µM,
2.0 mm × 30 mm, 5 µL injection volume) using a solvent
gradient of 10-100% acetonitrile in 0.1% aqueous trifluoro-
acetic acid over 4 min at a flow rate of 1.5 mL/min and
detecting by UV absorption at 220 and 254 nm and by ELSD.
X-r a y Cr ysta llogr a p h y. Crystals were obtained using
published conditions.9 As reported, the protein was found to
crystallize under a number of different conditions using the
commercially available Crystal Screen 1 (Hampton Research,
Laguna Niguel, CA) and was found to diffract to high resolu-
tion (1.6-2.0 Å). In this study, crystals from condition no. 26
(0.2 M ammonium acetate, 0.1 M trisodium citrate, pH 5.6,
30% MPD) were used for this series of experiments because
the mother liquor was found to be cryoprotectant. Crystal-
lographic data were processed, and structures were determined
and refined by standard methods using the program packages
DENZO,17 CNX, and QUANTA. Crystallographic screening
data were acquired using a Rigaku rotating anode generator
equipped with Osmic mirrors and a MarCCD detector. All
screening data were collected using a robotic crystal mounting
and alignment robot18 with data collection times of 2 h per
crystal. To facilitate rapid data collection, crystals were
visually prealigned in the mounting loops such that data would
be sampled along the crystallographic 4-fold axis. DHNA was
screened against a fragment library consisting of 10 000
compounds divided into 100 compound shape-diverse mixtures.
It was possible to achieve shape diversity for each mixture
because the relatively small compound library was designed
for maximum overall diversity to increase the likelihood of
finding lead compounds for a varied assortment of enzyme
active sites. The library and shape diverse mixtures were
generated computationally.10 The 10 000 compounds were
primarily obtained from the Abbott chemical repository,
although some were identified in the Available Chemicals
Directory (ACD) and purchased commercially.
En zym e In h ibition Assa y. Inhibition of DHNA activity
was quantified using a fluorescence-based kinetic enzyme
assay monitoring formation of the reaction product 6-hy-
droxymethyl-7,8-dihydropterin.19 The assay was performed in
a 96-well microtiter plate format using a Perkin-Elmer HTS
7000 Plus bioassay reader set for an excitation wavelength of
430 nm and an emission wavelength of 535 nm. Each reaction
was monitored for 1 h with the mean reaction rate integrated
over this time. The rate remained essentially linear throughout
this period when using the reaction parameters specified. Each
reaction was in a total volume of 150 µL in buffer consisting
of 20 mM HEPES (pH 7.4) and 150 mM NaCl. DHNA was
present at 50 nM concentration and showed no decrease in
activity following dilution or prolonged incubation. The sub-
strate 7,8-dihydroneopterin (Fluka, Alexis) was present at an
initial concentration of 4 µM. Reactions were run at ambient
temperature with the internal temperature of the plate reader
at approximately 29 °C.
3-Ca r boxyben zen ed ia zon iu m Tetr a flu or obor a te (21).14
3-Aminobenzoic acid (10.0 g, 72.9 mmol) was dissolved in 182
mL of ethanol and tetrafluoroboric acid (28.6 mL, 219 mmol)
was added, resulting in a clear solution with a small amount
of undissolved material. The mixture was cooled to 0 °C, and
isoamyl nitrite (29.4 mL, 219 mmol) was added dropwise. After
the mixture was stirred at 0 °C for 20 min, a precipitate
formed. Diethyl ether was added, and stirring continued for
30 min. The precipitate was filtered, rinsed thoroughly with
diethyl ether, and dried under vacuum. A total of 16.6 g (97%)
1
of white crystals was isolated. H NMR (300 MHz, DMSO-d6)
δ ppm 8.10 (t, J ) 7.8 Hz, 1H), 8.68 (ddd, J ) 8.2, 1.4, 1.1 Hz,
1H), 8.85 (ddd, J ) 8.2, 2.1, 1.1 Hz, 1H), 9.23 (t, J ) 1.7 Hz,
1H).
3-(2,4-Dia m in o-6-h yd r oxy-4,5-d ih yd r op yr im id in -5-yl-
a zo)ben zoic Acid (22).13 2,4-Diamino-6-hydroxypyrimidine
(8.35 g, 66.2 mmol) was dissolved in 500 mL of 1.4 M sodium
acetate, followed by slow addition of 3-carboxybenzenediazo-
nium tetrafluoroborate (16.4 g, 69.5 mmol). A thick, orange
precipitate quickly formed, and the mixture was stirred at
room temperature overnight. The precipitate was filtered,
rinsed with water, ethanol, and diethyl ether, and dried under
vacuum to afford 22.0 g (100%) of an orange solid consistent
with the trihydrate. 1H NMR (300 MHz, DMSO-d6) δ ppm 7.22
(br s, 2H), 7.48 (t, J ) 7.6 Hz, 1H), 7.79 (m, 2H), 7.92 (d, J )
3.7 Hz, 1H), 8.15 (s, 1H), 10.32 (br s, 1H), 11.08 (br s, 1H). MS
[(-)-ESI] m/z 273 [M - H]-.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)ben zoic Acid (19). Copper(II) sulfate (2.19 g, 13.7 mmol)
was dissolved in 30 mL of 1:1 water/pyridine, and the resulting
mixture was heated to reflux. Compound 22 (1.50 g, 4.57
mmol) was carefully added to the hot solution, causing the
reaction mixture to turn dark-green. After the mixture was
refluxed overnight, a light-blue suspension formed, which was
cooled to room temperature and poured into water. A precipi-
tate formed, which was filtered and rinsed with water and 5%
acetic acid. A green solid was isolated, which was dissolved in
5% sodium hydroxide and filtered. The yellow filtrate was
stirred over activated charcoal, filtered, and acidified with 1
M hydrochloric acid. A gelatinous precipitate formed, which
was filtered and rinsed thoroughly with water, followed by
ethanol and diethyl ether. A very light-green solid was isolated,
which was dried under vacuum to afford 1.24 g (100%) of the
title compound. 1H NMR (300 MHz, DMSO-d6) δ ppm 6.75 (br
s, 2H), 7.74 (t, J ) 8.1 Hz, 1H), 8.02 (ddd, J ) 7.9, 1.1, 1.0 Hz,
1H), 8.32 (ddd, J ) 8.1, 2.4, 1.0 Hz, 1H), 8.59 (t, J ) 2.0 Hz,
1H), 11.18 (s, 1H), 13.38 (br s, 1H). MS [(-)-ESI] m/z 271 [M
- H]-.
Gen er a l P r oced u r e for Am id e Syn th esis. Compounds
24-38 were synthesized in a LabTech Platform IV parallel
synthesizer from Advanced ChemTech using a 40-well solu-
tion-phase reaction block. In each well, compound 19 (0.200
g, 0.735 mmol), an appropriate amine (1.5 equiv, 1.10 mmol),
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride
(0.155 g, 0.808 mmol), 1-hydroxybenzotriazole hydrate (0.124
g, 0.808 mmol), and N,N-diisopropylethylamine (0.384 mL,
2.20 mmol) were combined in 3.7 mL of anhydrous, amine-
free N,N-dimethylformamide, and the mixtures were shaken
at 350 rpm at room temperature overnight. The reaction
mixtures were diluted with water and filtered, and the
resulting solids were rinsed sequentially with 1 M hydrochloric
acid, 5% sodium hydroxide, water, ethanol, and diethyl ether.
Syn th etic Ma ter ia ls a n d Meth od s. Unless otherwise
specified, reactions were performed in oven-dried glassware
under an inert atmosphere of nitrogen and monitored by thin-
layer chromatography (TLC) or by liquid chromatography-
mass spectrometry (LC-MS). All reagents were purchased
from commercial suppliers and used as provided. Analytical
TLC was performed on 0.25 mm precoated Merck silica gel 60
F
254, visualizing with ultraviolet light at 254, 302, or 365 nm
wavelength. Flash column chromatography was performed on
a Biotage Flash 40 system using 50 or 90 g of prepacked silica
columns and HPLC grade solvents. Analytical LC-MS was
performed on an Agilent series 1100 HPLC system equipped
with an autosampler and coupled to a Finnegan Thermoquest
atmospheric pressure chemical ionization (APCI) mass spec-
trometer. Peak detection was by ultraviolet absorption at 220
and 254 nm and by an evaporative light scattering detector
(ELSD). Proton nuclear magnetic resonance (1H NMR) spectra
were recorded on a Bruker ARX spectrometer (300 MHz) or
on a Varian Inova spectrometer (500 MHz) and are referenced
to residual solvent peaks (DMSO-d6, δ 2.50 ppm) or to an
internal standard of tetramethylsilane (TMS, δ 0.00 ppm). 1H-
1H couplings are assumed to be first-order, and peak multiplic-
ity is reported as s (singlet), d (doublet), t (triplet), q (quartet),
m (multiplet), and b (broad). Electrospray ionization (ESI) and

Inhibitors of Dihydroneopterin Aldolase
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7 1717
Each compound was isolated as an off-white powder requiring
no further purification.
NMR (300 MHz, DMSO-d6) δ ppm 0.85 (t, J ) 7.5 Hz, 3H),
1.29 (q, J ) 7.5 Hz, 2H), 3.25-3.33 (m, 6H), 4.50 (t, J ) 5.5
Hz, 2H), 6.70 (br s, 2H), 7.69 (m, 1H), 7.90 (d, J ) 7.7 Hz,
1H), 8.24 (d, J ) 7.7 Hz, 1H), 8.51 (s, 1H), 8.61 (t, J ) 5.7 Hz,
1H), 10.51 (br s, 1H). MS [(-)-ESI] m/z 386 [M - H]-.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(3-h yd r oxy-2,2-d im eth ylp r op yl)ben za m id e (34):
1H NMR (300 MHz, DMSO-d6) δ ppm 0.86 (s, 6H), 3.16 (d, J
) 6.1 Hz, 2H), 3.21 (d, J ) 6.1 Hz, 2H), 4.59 (t, J ) 5.8 Hz,
1H), 6.72 (br s, 2H), 7.70 (t, J ) 8.0 Hz, 1H), 7.94 (ddd, J )
8.0, 1.4, 1.2 Hz, 1H), 8.23 (ddd, J ) 8.2, 2.1, 0.85 Hz, 1H),
8.54 (t, J ) 1.9 Hz, 1H), 8.67 (t, J ) 6.3 Hz, 1H), 11.13 (s, 1H).
MS [(-)-ESI] m/z 356 [M - H]-. Anal. (C16H19N7O3‚0.5 H2O)
C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(1-h yd r oxycycloh exylm eth yl)ben za m id e (35): 1H
NMR (300 MHz, DMSO-d6) δ ppm 1.17-1.62 (m, 10H), 3.20-
3.30 (m, 2H), 4.38 (s, 1H), 6.75 (br s, 2H), 7.69 (t, J ) 8.0 Hz,
1H), 7.97 (d, J ) 7.8 Hz, 1H), 8.22 (dd, J ) 8.1, 1.4 Hz, 1H),
8.50 (t, J ) 5.6 Hz, 1H), 8.56 (s, 1H), 9.59 (br s, 1H). MS [(+)-
ESI] m/z 384 [M + H]+. Anal. (C18H21N7O3‚1.5 H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-cis-(2-h yd r oxycycloh exylm eth yl)ben za m id e (36):
1H NMR (300 MHz, DMSO-d6) δ ppm 1.20-1.25 (m, 1H), 1.28-
1.40 (m, 4H), 1.52-1.74 (m, 4H), 3.15-3.25 (m, 2H), 3.78 (br
s, 1H), 4.35 (d, J ) 4.1 Hz, 1H), 6.72 (br s, 2H), 7.69 (t, J ) 8.0
Hz, 1H), 7.94 (ddd, J ) 8.1, 1.4, 1.0 Hz, 1H), 8.22 (ddd, J )
8.1, 2.4, 1.0 Hz, 1H), 8.54 (t, J ) 1.9 Hz, 1H), 8.71 (t, J ) 5.8
Hz, 1H), 11.11 (br s, 1H). MS [(-)-ESI] m/z 382 [M - H]-.
Anal. (C18H21N7O3‚H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-t r a n s-(2-h yd r oxycycloh exylm et h yl)b en za m id e
(37): 1H NMR (300 MHz, DMSO-d6) δ ppm 0.91-1.22 (m, 4H),
1.40-1.87 (m, 5H), 3.10-3.19 (m, 1H), 3.23-3.34 (m, 1H), 3.53
(td, J ) 8.8, 4.8 Hz, 1H), 4.72 (d, J ) 5.1 Hz, 1H), 6.72 (br s,
2H), 7.69 (t, J ) 8.0 Hz, 1H), 7.94 (ddd, J ) 8.0, 1.2, 1.0 Hz,
1H), 8.22 (ddd, J ) 8.1, 2.2, 0.85 Hz, 1H), 8.54 (t, J ) 1.9 Hz,
1H), 8.67 (dd, J ) 6.1, 5.4 Hz, 1H), 11.12 (s, 1H). MS [(-)-
ESI] m/z 382 [M - H]-. Anal. (C18H21N7O3‚H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N -cis-(2-h yd r oxycycloh e p t ylm e t h yl)b e n za m id e
(38): 1H NMR (300 MHz, DMSO-d6) δ ppm 1.32-1.83 (m,
11H), 3.19-3.37 (m, 2H), 3.86 (br s, 1H), 4.37 (d, J ) 4.4 Hz,
1H), 6.73 (br s, 2H), 7.69 (t, J ) 8.0 Hz, 1H), 7.94 (ddd, J )
8.0, 1.4, 1.2 Hz, 1H), 8.23 (ddd, J ) 8.1, 2.0, 1.0 Hz, 1H), 8.55
(t, J ) 1.9 Hz, 1H), 8.72 (t, J ) 5.8 Hz, 1H), 11.13 (s, 1H). MS
[(+)-DCI] m/z 398 [M + H]+. Anal. (C19H23N7O3‚H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-bip h en yl-4-ylm eth ylben za m id e (24): 1H NMR (300
MHz, DMSO-d6) δ ppm 4.56 (d, J ) 5.8 Hz, 2H), 6.71 (br s,
2H), 7.35 (m, 1H), 7.41-7.49 (m, 4H), 7.60-7.66 (m, 4H), 7.72
(t, J ) 8.1 Hz, 1H), 8.02 (ddd, J ) 8.0, 1.4, 1.2 Hz, 1H), 8.25
(ddd, J ) 8.1, 2.4, 1.0 Hz, 1H), 8.63 (t, J ) 1.7 Hz, 1H), 9.40
(t, J ) 5.9 Hz, 1H), 11.11 (s, 1H). MS [(-)-ESI] m/z 496 [M +
CH3CO2]-. Anal. (C24H19N7O2) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(3,5-bistr iflu or om eth ylben zyl)ben za m id e (25): 1H
NMR (300 MHz, DMSO-d6) δ ppm 4.70 (d, J ) 5.8 Hz, 2H),
6.36 (br s, 2H), 7.68 (t, J ) 8.0 Hz, 1H), 7.92 (d, J ) 8.1 Hz,
1H), 8.01 (s, 1H), 8.06 (s, 2H), 8.23 (ddd, J ) 8.1, 2.0, 0.68 Hz,
1H), 8.59 (s, 1H), 9.48 (t, J ) 5.8 Hz, 1H). MS [(-)-ESI] m/z
496 [M - H]-.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
1
2-yl)-N-(4-p h en oxyben zyl)ben za m id e (26): H NMR (300
MHz, DMSO-d6) δ ppm 4.50 (d, J ) 5.9 Hz, 2H), 6.74 (br s,
2H), 6.99 (m, 4H), 7.12 (t, J ) 7.4 Hz, 1H), 7.38 (m, 4H), 7.70
(t, J ) 7.9 Hz, 1H), 8.00 (d, J ) 8.1 Hz, 1H), 8.24 (dt, J ) 8.1,
1.1 Hz, 1H), 8.61 (t, J ) 1.7 Hz, 1H), 9.37 (t, J ) 5.9 Hz, 1H),
11.20 (br s, 1H). MS [(-)-ESI] m/z 452 [M - H]-. Anal.
(C24H19N7O3‚H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(3,5-d ich lor oben zyl)ben za m id e (27): 1H NMR (300
MHz, DMSO-d6) δ ppm 4.52 (d, J ) 5.9 Hz, 1H), 6.73 (br s,
2H), 7.40 (d, J ) 1.8 Hz, 2H), 7.51 (t, J ) 1.8 Hz, 1H), 7.72 (t,
J ) 8.1 Hz, 1H), 8.01 (dt, J ) 8.1, 1.5 Hz, 1H), 8.26 (ddd, J )
8.1, 2.2, 1.1 Hz, 1H), 8.61 (t, J ) 1.8 Hz, 1H), 9.42 (t, J ) 5.9
Hz, 1H), 11.11 (s, 1H). MS [(-)-ESI] m/z 428 [M - H]-. Anal.
(C18H13Cl2N7O2‚0.5 H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-ben zo[1,3]d ioxol-5-ylm eth ylben za m id e (28): 1H
NMR (300 MHz, DMSO-d6) δ ppm 4.41 (d, J ) 5.9 Hz, 1H),
5.98 (s, 2H), 6.73 (br s, 2H), 6.81-6.92 (m, 3H), 6.97-7.01 (m,
1H), 7.35-7.40 (m, 1H), 7.69 (t, J ) 7.9 Hz, 1H), 7.97 (d, J )
7.7 Hz, 1H), 8.23 (dd, J ) 7.7, 1.8 Hz, 1H), 8.59 (t, J ) 1.8 Hz,
1H), 9.29 (t, J ) 5.9 Hz, 1H), 11.05 (br s, 1H). MS [(-)-ESI]
m/z 404 [M - H]-.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N -[2-(2-h yd r oxym e t h ylp h e n ylsu lfa n yl)b e n zyl]-
1
ben za m id e (29): H NMR (300 MHz, DMSO-d6) δ ppm 4.59
(t, J ) 6.4 Hz, 4H), 5.33 (t, J ) 5.5 Hz, 1H), 6.73 (br s, 2H),
7.06-7.15 (m, 2H), 7.21-7.29 (m, 2H), 7.31-7.41 (m, 3H), 7.57
(d, J ) 7.4 Hz, 1H), 7.71 (t, J ) 8.1 Hz, 1H), 8.00 (d, J ) 8.1
Hz, 1H), 8.25 (dd, J ) 8.8, 1.5 Hz, 1H), 8.63 (s, 1H), 9.32 (t, J
) 5.7 Hz, 1H), 11.13 (m, 1 H). MS [(-)-ESI] m/z 498 [M -
H]-. Anal. (C25H21N7O3‚0.5 H2O) C, H, N.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra for
compounds 19, 21, 22, and 24-38. X-ray parameters and
Protein Data Bank file names for DHNA/inhibitor complexes
of compounds 11-15, 19, 27, and 29. This material is available
free of charge via the Internet at http://pubs.acs.org.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(2,3-d ih yd r ob en zofu r a n -5-ylm et h yl)b en za m id e
(30): 1H NMR (300 MHz, DMSO-d6) δ ppm 3.15 (t, J ) 8.7
Hz, 2H), 4.41 (d, J ) 6.1 Hz, 2H), 4.49 (t, J ) 8.8 Hz, 2H),
5.40 (s, 2H), 6.70 (d, J ) 8.1 Hz, 1H), 7.07 (dd, J ) 8.1, 2.0
Hz, 1H), 7.22 (s, 1H), 7.61 (t, J ) 8.0 Hz, 1H), 7.83 (ddd, J )
7.9, 1.3, 1.0 Hz, 1H), 8.15 (ddd, J ) 8.1, 2.4, 1.0 Hz, 1H), 8.53
(t, J ) 1.9 Hz, 1H), 9.22 (t, J ) 5.8 Hz, 1H). MS [(+)-ESI] m/z
402 [M - H]-. Anal. (C20H17N7O3‚H2O) C, H, N.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(3-h yd r oxyp r op yl)ben za m id e (31): 1H NMR (500
MHz, DMSO-d6) δ ppm 1.66 (m, 2H), 3.30 (t, J ) 6.1 Hz, 2H),
3.37-3.46 (m, 3H), 6.68 (br s, 2H), 7.63 (t, J ) 8.2 Hz, 1H),
7.89 (d, J ) 7.9 Hz, 1H), 8.17 (d, J ) 7.9 Hz, 1H), 8.50 (s, 1H),
8.69 (t, J ) 5.5 Hz, 1H). MS [(-)-ESI] m/z 328 [M - H]-.
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(4-h yd r oxybu tyl)ben za m id e (32): 1H NMR (300
MHz, DMSO-d6) δ ppm 1.43-1.63 (m, 4H), 3.40-3.50 (m, 4H),
4.40 (t, J ) 5.2 Hz, 1H), 6.74 (br s, 2H), 7.68 (t, J ) 7.9 Hz,
1H), 7.93 (d, J ) 8.1 Hz, 1H), 8.21 (dd, J ) 7.7, 1.8 Hz, 1H),
8.55 (t, J ) 1.8 Hz, 1H), 8.77 (t, J ) 5.5 Hz, 1H), 11.19 (br s,
1H). MS [(-)-ESI] m/z 342 [M - H]-.
Refer en ces
(1) Chopra, I.; Hesse, L.; O’Neill, A. J . Exploiting current under-
standing of antibiotic action for discovery of new drugs. Soc.
Appl. Microbiol. Symp. Ser. 2002, 31, 4S-15S. Daw, M. A.; Drah,
A. M. Antibiotic resistance: Prospects for the new millennium.
J . Chemother. 2001, 13, 587-594. Bax, R.; Mullan, N.; Verhoef,
J . The millennium bugssthe need for and development of new
antibacterials. Int. J . Antimicrob. Agents 2000, 16, 51-59.
Domagala, J . M.; Sanchez, J . P. New approaches and agents to
overcome bacterial resistance. Annu. Rep. Med. Chem. 1997, 32,
111-120.
(2) Lerner, C. G.; Beutel, B. A. Antibacterial drug discovery in the
post-genomics era. Curr. Drug Targets: Infect. Disord. 2002, 2,
109-119. Chan, P. F.; Macarron, R.; Payne, D. J .; Zalacain, M.;
Holmes, D. J . Novel antibacterials: a genomics approach to drug
discovery. Curr. Drug Targets: Infect. Disord. 2002, 2, 291-308.
Volker, C.; Brown, J . R. Bioinformatics and the discovery of novel
anti-microbial targets. Curr. Drug Targets: Infect. Disord. 2002,
2, 279-290. Chalker, A. F.; Lunsford, R. D. Rational identifica-
tion of new antibacterial drug targets that are essential for
viability using a genomics-based approach. Pharmacol. Ther.
2002, 95, 1-20. Isaacson, R. E. Genomics and the prospects for
the discovery of new targets for antibacterial and antifungal
agents. Curr. Pharm. Des. 2002, 8, 1091-1098. McDevitt, D.;
3-(5-Am in o-7-h yd r oxy-[1,2,3]tr ia zolo[4,5-d ]p yr im id in -
2-yl)-N-(2,2-bis-h yd r oxym eth ylbu tyl)ben za m id e (33): 1H

1718 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 7
Sanders et al.
Rosenberg, M. Exploiting genomics to discover new antibiotics.
Trends Microbiol. 2001, 9, 611-617. Payne, D. J .; Holmes, D.
J .; Rosenberg, M. Delivering novel targets and antibiotics from
genomics. Curr. Opin. Invest. Drugs 2001, 2, 1028-1034.
Holmes, D. J .; Throup, J . P.; Wallis, N. G.; Burnham, M. K. R.;
Zalacain, M.; Biswas, S.; Chalker, A. F.; Ingraham, K. A.; Marra,
A.; Bryant, A.; Woodnutt, G.; Warren, P. V.; Brown, J . R.;
Rosenberg, M. The contribution of genomics to the discovery of
new antibiotics. Focus Biotechnol. 2001, 1, 23-31. Lehoux, D.
E.; Levesque, R. C. Detection of genes essential in specific niches
by signature-tagged mutagenesis. Curr. Opin. Biotechnol. 2000,
11, 434-439. Reich, K. A. The search for essential genes. Res.
Microbiol. 2000, 151, 319-324. Gray, C. P.; Keck, W. Bacterial
targets and antibiotics. Genome-based drug discovery. Cell. Mol.
Life Sci. 1999, 56, 779-787. Chopra, I. Prospects for antisense
agents in the therapy of bacterial infections. Expert Opin. Invest.
Drugs 1999, 8, 1203-1208. Kolstoe, A.-B. Dynamic bacterial
genome organization. Mol. Microbiol. 1997, 24, 241-248. Miklos,
G. L. G.; Rubin, G. M. The role of the genome project in
determining gene function: insights from model organisms. Cell
1996, 86, 521-529.
Methods Enzymol. 1971, 18, 779-786. Hitchings, G. H.; Bur-
chall, J . J .; Ferone, R. Comparative enzymology of dihydrofolate
reductase and the design of chemotherapeutic agents. Symp. Soc.
Gen. Microbiol. 1966, 16, 294-300.
(7) Mathis, J . B.; Brown, G. M. Biosynthesis of folic acid. XI.
Purification and properties of hydroneopterin aldolase. J . Biol.
Chem. 1970, 245, 3015-3025.
(8) Zimmerman, M.; Tolman, R. L.; Morman, H.; Graham, D. W.;
Rogers, E. F. Inhibitors of folate biosynthesis. 1. Inhibition of
dihydroneopterin aldolase by pteridine derivatives. J . Med.
Chem. 1977, 20, 1213-1215.
(9) Hennig, M.; D’Arcy, A.; Hampele, I. C.; Page, M. G. P.; Oefner,
C.; Dale, G. E. Crystal structure and reaction mechanism of 7,8-
dihydroneopterin aldolase from Staphylococcus aureus. Nat.
Struct. Biol. 1998, 5, 357-362.
(10) Nienaber, V. L.; Richardson, P. L.; Klighofer, V.; Bouska, J . J .;
Giranda, V. L.; Greer, J . Discovering novel ligands for macro-
molecules using X-ray crystallographic screening. Nat. Biotech-
nol. 2000, 18, 1105-1108.
(11) Muchmore, S. W. Experiences with CCD detectors on a home
X-ray source. Acta Crystallogr., Sect. D: Biol. Crystallogr. 1999,
D55, 1669-1671.
(12) Bode, W.; Schwager, P. Refined crystal structure of bovine.beta.-
trypsin at 1.8.ANG. resolution. II. Crystallographic refinement,
calcium binding site, benzamidine binding site, and active site
at pH 7.0. J . Mol. Biol. 1975, 98, 693-717.
(13) Benson, F. R.; Hartzel, L. W.; Savell, W. L. v-Triazolo[d]-
pyrimidines. I. 2-Aryl-5-amino-7-hydroxy derivatives. J . Am.
Chem. Soc. 1950, 72, 1816-1818.
(14) Crich, D.; Hao, X. Generation and Cyclization of Acyl Radicals
from Thiol Esters under Nonreducing, Tin-Free Conditions. J .
Org. Chem. 1997, 62, 5982-5988.
(3) Bermingham, A.; Derrick, J . P. The folic acid biosynthesis
pathway in bacteria: Evaluation of potential for antibacterial
drug discovery. BioEssays 2002, 24, 637-648. Miller, G. H.;
Doukas, P. H.; Seydel, J . K. Sulfonamide structure-activity
relation in a cell-free system. Correlation of inhibition of folate
synthesis with antibacterial activity and physicochemical pa-
rameters. J . Med. Chem. 1972, 15, 700-706. Hitchings, G. H.
Folate antagonists as antibacterial and antiprotozoal agents.
Ann. N. Y. Acad. Sci. 1971, 186, 444-451. Hitchings, G. H.
Inhibitors of dihydrofolic acid reductase in the therapy of
infectious diseases. Bibl. Nutr. Diet. 1966, 8, 226-237.
(4) Brown, G. M. Biosynthesis of pterins. Folates Pterins 1985, 2,
115-154.
(5) McCullough, J . L.; Maren, T. H. Inhibition of dihydropteroate
synthetase from Escherichia coli by sulfones ad sulfonamides.
Antimicrob. Agents Chemother. 1973, 3, 665-669. Ho, R. I.;
Corman, L.; Morse, S. A.; Schneider, H. Structure activity of
sulfones and sulfonamides on dihydropteroate synthetase from
Neisseria meningitidis. Antimicrob. Agents Chemother. 1975, 7,
758-763. Vinnicombe, H. G.; Derrick, J . P. Dihydropteroate
synthase: an old drug target revisited. Biochem. Soc. Trans.
1999, 27, 53-58.
(15) Herrington, M. B. Measurement of the uptake of radioactive
para-aminobenzoic acid monitors folate biosynthesis in Escheri-
chia coli K-12. Anal. Biochem. 1994, 216, 427-430.
(16) Quinlivan, E. P.; McPartlin, J .; Weir, D. G.; Scott, J . Mechanism
of the antimicrobial drug trimethoprim revisited. FASEB J .
2000, 14, 2519-2524.
(17) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data
collected in oscillation mode. Methods Enzymol. 1997, 276, 307-
326.
(18) Muchmore, S. W.; Olson, J .; J ones, R.; Pan, J .; Blum, M.; Greer,
J .; Merrick, S. M.; Magdalinos, P.; Nienaber, V. L. Automated
Crystal Mounting and Data Collection for Protein Crystal-
lography. Structure 2000, 8, R243-R246.
(6) Hitchings, G. H., J r. Selective inhibitors of dihydrofolate reduc-
tase (Nobel report). Angew. Chem. 1989, 101, 903-909. Periti,
P. Evolution of the bacterial dihydrofolate reductase inhibitors.
J . Antimicrob. Chemother. 1995, 36, 887-890. Roth, B.; Bur-
chall, J . J . Small molecule inhibitors of dihydrofolate reductase.
(19) Harlan, J . Manuscript in preparation.
J M030497Y