Design of new benzoxazole-2-thione-derived inhibitors of Streptococcus pneumoniae hyaluronan lyase: structure of a complex with a 2-phenylindole
Design of new benzoxazole-2-thione-derived inhibitors of Streptococcus pneumoniae hyaluronan lyase: structure of a complex with a 2-phenylindole
Daniel J. Rigden 0 2
Alexander Botzki 1
Masatoshi Nukui 0
R. Brandon Mewbourne 0
Ejvis Lamani 0
Stephan Braun 1
Erwin von Angerer 1
Günther Bernhardt 1
Stefan Dove 1
Armin Buschauer 1
Mark J. Jedrzejas 0
0 Children's Hospital Oakland Research Institute , Oakland, CA 94609
1 Institute of Pharmacy, University of Regensburg , 93040 Regensburg , Germany
2 School of Biological Sciences, University of Liverpool , Crown Street, Liverpool L69 7ZB , UK
The bacterial hyaluronan lyases (Hyals) that degrade hyaluronan, an important component of the extracellular matrix, are involved in microbial spread. Inhibitors of these enzymes are essential in investigation of the role of hyaluronan and Hyal in bacterial infections and constitute a new class of antibiotics against Hyal-producing bacteria. Recently, we identified 1,3diacetylbenzimidazole-2-thione and related molecules as inhibitors of streptococcal Hyal. One of such compounds, 1-decyl-2(4-sulfamoyloxyphenyl)-1-indol-6-yl sulfamate, was co-crystallized in a complex with Streptococcus pneumoniae Hyal and its structure elucidated. The resultant X-ray structure demonstrates that this inhibitor fits in the enzymatic active site via interactions resembling the binding mode of the natural hyaluronan substrate. X-ray structural analysis also indicates binding interactions with the catalytic residues and those of a catalytically essential hydrophobic patch. An IC50 value of 11 mM for Hyal from Streptococcus agalactiae (strain 4755) qualifies this phenylindole compound as one of the most potent Hyal inhibitors known to date. The structural data suggested a similar binding mode for N-(3-phenylpropionyl)-benzoxazole-2-thione. This new compound's inhibitory properties were confirmed resulting in discovery of yet another Hyal inhibitor (IC50 of 15 mM). These benzoxazole-2-thiones constitute a new class of inhibitors of bacterial Hyals and are well suited for further optimization of their selectivity, potency, and pharmacokinetic properties.
antimicrobials/hyaluronidase/molecular mechanism/pneumococci/spreading factor
Introduction The glycosaminoglycan hyaluronan (hyaluronic acid [HA]) was first isolated from the vitreous humor of bovine eyes
*These authors contributed equally to this work.
1To whom correspondence should be addressed; e-mail:
2To whom correspondence should be addressed; e-mail:
(Meyer and Palmer, 1934) and is a major component of the
extracellular matrix (ECM) of animal tissues. This strictly
linear polymer consists of repeating disaccharide units of
It has been identified in the capsule of some bacteria and in
essentially all vertebrates where it is present in significant
quantities in skin (dermis and epidermis), brain, and the
central nervous system (Laurent and Fraser, 1992). HA is
degraded by hyaluronidases, a class of enzymes with two
main families, the eukaryotic hyaluronoglucosaminidases
(EC 126.96.36.199) and the prokaryotic hyaluronan lyases
(Hyals) (EC 188.8.131.52) (Kreil, 1995; Coutinho and Henrissat,
The best characterized of the bacterial enzymes are
Streptococcus pneumoniae and Streptococcus agalactiae
Hyals (Jedrzejas and Chantalat, 2000; Jedrzejas, 2002b).
Both enzymes degrade HA at the β-1,4-glycosidic linkage
between D-glucuronic acid and N-acetyl-D-glucosamine.
The product of this elimination reaction is the unsaturated
2-acetamido-2-deoxy-3-O-(β-D-gluco-4-enepyranosyluronic acid) D-glucose (Kreil, 1995; Linker and
Meyer, 1954). The three-dimensional (3D) structures of
S. pneumoniae and S. agalactiae strain 3502 Hyals were
elucidated by X-ray analyses in the native as well as in the
complex forms with both substrates and products (Jedrzejas
and Chantalat, 2000; Li and Jedrzejas, 2001; Li, et al., 2000;
Jedrzejas, 2002b). The structures of both enzymes are
similar and contain distinct domains with the largest domain
being catalytic. The active site of the S. pneumoniae enzyme
(SpnHyal), schematically depicted in Figure 1, is located in
a long cleft of the catalytic domain that binds the substrate
and consists of four main parts: a positive patch, a catalytic
triad (His399, Tyr408, and Asn349) responsible for
degradation, a hydrophobic/aromatic patch (Trp291, Trp292,
and Phe343) responsible for correct positioning of the
substrate chains for catalysis, and a negative patch implicated
in the release of the product (Glu388, Asp398, and Thr400)
(Li et al., 2000). Successive structural (Li et al., 2000;
Ponnuraj and Jedrzejas, 2000; Li and Jedrzejas, 2001;
Jedrzejas et al., 2002; Mello et al., 2002) and mutagenesis
studies (Li et al., 2000; Kelly et al., 2001; Nukui et al., 2003)
have enabled the formulation of a firmly supported catalytic
mechanism, as reported (Jedrzejas, 2001, 2002a,b). This
mechanism is based on acid/base type β-elimination with
His399 serving as base and Tyr408 as acid (Figure 1).
The Hyal enzymes identified in two S. agalactiae strains,
4755 (SagHyal4755) and 3502 (SagHyal3502), have nearly
identical sequences (98% identity) (Ponnuraj and Jedrzejas,
2000) and show 53% identity with the S. pneumoniae
enzyme. In contrast to the structure of SpnHyal, a fully
active two-domain truncated version (Li et al., 2000), the
Fig. 1. Binding of a hyaluronan hexasaccharide in the catalytic cleft of
Streptococcus pneumoniae hyaluronan lyase. HA with disaccharide units
HA1 (reducing end), HA2 and HA3 (non-reducing end), and the catalytic
triad (Tyr408, His399, and Asn349) were extracted from the SpnHyal–
hexasaccharide HA complex (Jedrzejas et al., 2002). Oxygen atoms are
drawn as black balls and nitrogen as smaller grey balls. The positions of
the “patches” forming the binding site with respect to the substrate are
indicated. Dotted lines indicate interactions central to the catalytic
mechanism according to Li and others (2000).
SagHyal3502 structure reveals an additional third domain at
the N-terminus, which is presumed to act as a spacer (Li
and Jedrzejas, 2001). Bioinformatic studies identified a
further domain at the extreme N-terminus in all bacterial Hyal
enzymes, including S. pneumoniae and S. agalactiae, that is
responsible for substrate binding but not for catalysis
(Rigden and Jedrzejas, 2003).
Unsaturated disaccharides are the only degradation
products of exhaustive digestion of HA by Hyals. These
disaccharides might be utilized by bacteria as an additional
carbon source (Hynes and Walton, 2000). The degradation
of HA in the host ECM decreases the viscoelasticity of the
ground substance, which presumably facilitates bacterial
spread and their associated toxins into tissues (Hynes and
Walton, 2000). Streptococcal Hyal is therefore a significant
virulence factor. Human infection by such pathogens is one
of the major causes of meningitis, septicemia, as well as
many other serious diseases that lead to neonatal mortality
(Dillon et al., 1987; Hynes and Walton, 2000). To study the
precise role of HA and Hyals in the course of bacterial
infection and spread, we undertook the design and
development of inhibitors in our laboratories.
For bacterial Hyals, L-ascorbic acid (Vc, 1, Figure 2) was
described as a competitive inhibitor with an IC50 value of
∼6 mM (Li et al., 2001). Furthermore, various unsaturated
fatty acids have been reported to inhibit S. dysgalactiae
Hyal, with IC50 values ranging from 1 to 172 μM (Suzuki
et al., 2002). The structural basis of the SpnHyal–Vc
interaction was elucidated by means of X-ray crystallography
(Li et al., 2001). Vc binds in the active site of the enzyme,
and among the different interactions, the importance of
hydrophobic contacts with Trp292 and Tyr408 was evident.
On the basis of the hypothesis that Vc derivatives with
increased hydrophobicity would be more potent, L-ascorbic
acid-6-hexadecanoate (Vc palmitoate, Vcpal, 2) (Figure 2),
a highly effective antioxidant (Klein and Weber, 2001) and
Fig. 2. Chemical structures of hyaluronan lyase inhibitors. 1: Vitamin C,
L-ascorbic acid, Vc; 2: L-ascorbic acid-6-hexadecanoate, Vcpal; 3:
1-decyl-2-(4-sulfamoyloxyphenyl)-1Hindol-6-yl sulfamate; and 5: N-substituted benzoxazole-2-thione
glutathione-S-transferase inhibitor (Mitra et al., 1992), was
examined (Botzki et al., 2004). The X-ray structure of the
SpnHyal-Vcpal complex [Protein Data Bank (PDB) (Berman
et al., 2000) code 1w3y (Botzki et al., 2004)], determined at
1.65 Å resolution, confirmed the hypothesis that additional
hydrophobic interactions with Phe343, His399, and Thr400
in the active site lead to increased inhibition.
On the basis of these studies of bacterial Hyals inhibition
by Vc and Vcpal and the discovery that SagHyal is
inhibited by, for example, 1,3-diacetylbenzimidazole-2-thione (3,
Figure 2), a new inhibitor was designed,
1-decyl-2-(4-sulfamoyloxyphenyl)-1H-indol-6-yl sulfamate (4, Figure 2)
(Walter et al., 2004). In this study, the X-ray structure of a
complex of SpnHyal with this new inhibitor, compound 4,
is presented and its binding mode compared with that of a
hexasaccharide HA substrate and Vcpal. In addition, the
newly identified interaction pattern of compound 4 was
used to predict N-substituted benzoxazole-2-thiones as
additional novel Hyal inhibitors. The initial results on
structure-activity relationships (SAR) in this new series
support the binding mode derived from the crystallized
complexes and provide a basis for further optimization of
the molecular structures of these compounds with respect to
their selectivity, potency, and pharmacokinetic properties.
Results and Discussion
Initial design of inhibitors
Recent structure-based design approaches using the
program LUDI (Böhm, 1992) for docking of molecules from
databases into a SagHyal4755 model led to the prediction of
numerous potential inhibitors (Botzki et al., 2005).
Nineteen compounds were synthesized, from which 13 proved to
be active at SagHyal4755 (Botzki, et al., 2005).
1,3-diacetylbenzimidazole-2-thione (3, Figure 2) was the most potent
inhibitor (IC50 value of 5 μM at pH 7.4). Another hit from
these investigations, indole-2-carboxylic acid, was active only
at concentrations above 1 mM. Optimization of this structure
resulted in a series of inhibitors based on the 2-phenylindole
scaffold (Salmen, 2003; Walter et al., 2004). The hypothesis
that hydrophobic chains at the indole nitrogen may mimic the
hexadecyl moiety of Vcpal (2, Figure 2) was proposed and
later confirmed. Furthermore, the most active compound of
the 2-phenylindole series,
1-decyl-2-(4-sulfamoyloxyphenyl)1H-indol-6-yl sulfamate (4, Figure 2), was crystallized in
complex with SpnHyal.
Structure of the S. pneumoniae Hyal–phenylindole inhibitor
The overall protein structure of the SpnHyal-phenylindole
inhibitor complex (Table I) is similar to that of the
twodomain arrangement in the original native apo-form
structure of this enzyme (Jedrzejas et al., 1998; Li et al., 2000,
Ponnuraj and Jedrzejas, 2000). The crystallized Hyal molecule
is composed of two domains, the catalytic domain having
an α5/α5 barrel structure (α-domain) and the other
Low resolution diffraction limit (Å)
High resolution diffraction limit (Å)
Non-hydrogen xylitol atoms
Non-hydrogen inhibitor atoms
Non-hydrogen solvent atoms
Number of reflectionsa
Mean temperature factor B (Å2)
Protein main chain
Protein side chain
Bond lengths (Å)
Bond angles (°)
rms deviation from ideal values
comprised mainly of an antiparallel, three-layer β-sandwich
(β-domain). The catalytic cleft transverses the catalytic
αdomain. The active site of the enzyme is located within the
barrel domain on one side of the cleft area (Li and Jedrzejas,
2001; Jedrzejas, 2002a,b; Jedrzejas et al., 2002; Mello et al.,
2002). The β-domain is located at the C-terminus and does
not participate in catalysis. It is presumed to modulate the
substrate access to the cleft by covering the cleft area of the
α-domain because of its inherently great flexibility and
motion (Jedrzejas, 2002; Mello et al., 2002). The full-length
enzyme contains additional residues at the N-terminus
arranged in two domains (Rigden and Jedrzejas, 2003) that
are not catalytically important. The first of these domains is
a substrate-binding domain that is thought to act to
enhance the Hyal affinity for HA (for full discussion see
(Rigden and Jedrzejas, 2003; Rigden et al., 2003). The
second, smaller domain acts as a spacer to distance the
catalytic domain from the N-terminal substrate-binding
domain (Rigden and Jedrzejas, 2003).
The inhibitor 4 was found to bind in the cleft present in
the α-domain in the area of the catalytic and hydrophobic
patch residues (Figure 1). Inhibitor binding was clearly
evident from the extra electron density in the active site area,
which corresponded to the size and shape of the compound.
In general, this new inhibitor is located in the same location
in which the HA disaccharide product (HA1 in Figure 1)
(Ponnuraj and Jedrzejas, 2000), Vc (Li et al., 2001) as well
as Vcpal (Botzki et al., 2004) were found to bind in our
earlier studies. The location of this inhibitor in the active site
area suggests its competitive nature and its ability to
prevent the catalytic function of this enzyme.
The binding site of the phenylindole inhibitor
Electron density at the catalytic site allowed for the
satisfactory modeling of the phenylindole as shown in Figure 3.
The inhibitor was located within the electron density with
the exception of the sulfamoyl group in 4-position of the
phenyl moiety. This presumably reflects this sulfamoyl
group solvent exposure and suggests free rotation about the
NH2SO2O–phenyl bond. As refinement progressed, it
became clear that the 6-sulfamoyloxy group at the indole
moiety may adopt two distinct conformations (Figure 3).
These alternate binding modes are associated with a small,
nearly parallel displacement of the indole ring. The sulfamoyl
aValues in brackets are for the highest resolution shell, 2.07–2.00 Å.
bEach of the two modeled alternate inhibitor-binding modes comprises
Fig. 3. Stereo view of the SpnHyal cleft with the bound phenylindole
inhibitor 4. Density supports the existence of two modes of interaction
(modes 1 and 2). The sigmaA-weighted |2Fo-Fc| map electron density is
shown at a contouring level of 0.65σ.
position in mode 1 is well defined by density. In mode 2, it
is not as clearly defined but, if omitted, leads to the
presence of significant positive difference electron density.
Thus, its inclusion in the model was necessary. It appears
that mode 1, shown in more detail in Figure 4, is favored
for energetic reasons, but at the resolution available,
further occupation refinement is questionable. In the final
model, additional difference density was evident at the
catalytic site, suggestive of the existence of yet additional
modes of inhibitor binding. However, only the
conformation shown was well supported by density.
Fig. 4. SpnHyal-binding site of the phenylindole inhibitor 4. The
orientation shown faces into the substrate-binding cleft with the α-domain
toward the top of the page and the β-domain toward the bottom.
(A) Detailed representation of the interactions. Carbon atoms are colored
as follows—inhibitor white, residues with hydrophobic contacts to the
inhibitor orange, other residues green. Red spheres represent bound water
molecules and yellow dashed lines indicate hydrogen bonds. (B)
Lipophilic potential of the binding site mapped onto a MOLCAD surface.
Hydrophobic regions are colored in brown, polar regions in green and
green-blue. The 4-sulfamoyloxy substituent is exposed to the solvent.
The binding of the phenylindole 4 to SpnHyal involves
both hydrophobic interactions of the rings and the aliphatic
tail and hydrogen bonds of the 6-sulfamoyloxy group. The
indole plane of the inhibitor is in parallel with that of
Trp292 (Figures 3 and 4A) in an energetically favorable
distance of ca. 3.5 Å. The other side of the indole moiety
interacts with the guanidino group of Arg462. Such cation-π
interactions frequently contribute to the free energy of
protein-ligand binding (Gallivan and Dougherty, 1999).
The phenyl moiety of the inhibitor is sandwiched between
Trp291 and Asn580. The interaction with Trp291 involves
an approximately perpendicular ring arrangement (Figure 3).
The aliphatic tail fits into a surface crevice lined mainly by
hydrophobic residues—Met579, Trp291, and Phe343—and
also by Arg336, Glu388, and His399. The lipophilic
potential mapped on the surface of the binding site depicts a
continuous hydrophobic region extending from Trp292
outward to Phe343, which aligns with the indole ring and
nearly the complete decyl chain (Figure 4B).
The 6-sulfamoyloxy group forms a network of hydrogen
bonds with protein and water. The interactions associated
with binding mode 1 (Figure 4A) seem more favorable,
with charge-assisted bidentate H bonds between Arg466
and two sulfamoyl oxygen atoms. Both the oxygen atoms
and nitrogen atoms are additionally linked to the protein
by through–water interactions. In binding mode 2, there is
only a single direct H bond of a sulfamoyl oxygen with the
indole NH of Trp292. In this case, a water molecule
mediates additional interactions with the carboxylate of Asp352.
The binding site of the phenylindole inhibitor overlaps
with that of the HA substrate (see also Figure 1). Figure 5A
presents an alignment of the structure of functional
SpnHyal-hexasaccharide HA complex (Jedrzejas et al.,
2002) (substrate-based hexasaccharide bound to SpnHyal,
PDB code 1loh) with the SpnHyal-inhibitor complex (root
mean square [rms] of the backbone atoms 0.37 Å). The
amino acids of the catalytic triad, Asn349, His399, and
Tyr408, are involved in interactions with the phenylindole.
A water-mediated hydrogen bond (Figure 4A) links the
amide NH2 of Asn349 and the oxygen in 6-position of the
indole in binding mode 1, the imidazole ring of His399
contributes to the fit of the decyl chain, and the phenyl moiety
of Tyr408 is approached by the NH2 function of the
6-sulfamoyloxy group. Also, the hydrophobic patch (Trp291, Trp292,
and Phe343) that participates in the precise positioning of
the substrate belongs to the binding site of the inhibitor.
Comparing the binding modes of the HA substrate and the
inhibitor, it is obvious that the indole moiety of compound 4
is essentially coplanar with HA and is located in the
HA-binding region around the scissile 1,4-glycosidic bond. The
Nacetyl group of HA is directly related to the 6-sulfamoyloxy
substituent of compound 4. The N-acetyl-D-glucosamine unit
of HA approximately corresponds to the decyl chain.
Comparison of the binding modes of the phenylindole
Apart from the alkyl chains, the phenylindole 4 and Vcpal
are prima facie structurally dissimilar (Figure 2).
Nevertheless, their bound conformations and the binding modes are
similar to some extent and are therefore compared in some
detail. X-ray crystallography of the co-crystallized SpnHyal–
Vcpal complex resulted in the PDB structure 1w3y (Botzki
et al., 2004). Figure 5B presents a superposition of 1w3y
onto the complex of SpnHyal with compound 4 (rms of the
backbone atoms 0.07 Å). The vitamin C moiety, as
illustrated by the structure of SpnHyal–Vcpal complex, binds in
a ring-opened form, and it occupies the same region as
intact Vc (Li et al., 2001). This Vcpal ring opening also
occurs during the oxidation of Vc into L-dehydro-ascorbic
acid which is rapidly degraded under physiological
conditions (pH 7.4) via L-diketo-gulonate to yield L-erythrulose
and oxalate (Simpson and Ortwerth, 2000). The L-ascorbic
acid moiety of Vcpal binds within the catalytic site
overlapping with the indole ring of 4. Most protein–Vcpal
interactions are hydrophobic. The hydrophobic face of the Vc
portion is stacked with the indole ring of Trp292. This
structural arrangement helps the Vc moiety of Vcpal, like
Vc itself, to bind in the narrowest part of the binding site
comprising the amino acids Trp292, Arg462, and Arg466
(Li et al., 2001). The Vcpal palmitoyl moiety fits
approximately in the same hydrophobic surface crevice as the decyl
chain of the phenylindole. The binding of the proximal
aliphatic tails of Vcpal and the phenylindole is generally
similar, starting from overlapping positions of the palmitoyl C1
and the indole N1 atoms (distance ∼1 Å). A significant
difference is evident at the distal ends of the tail chains. The
shorter chain length in compound 4 (10 carbons, as
opposed to the 16 of Vcpal) enables the distal methylene
groups to fit into a surface pocket of the enzyme, in
contrast to the solvent exposure of the end of the palmitoyl
moiety. This fit comes along with closer binding of the
terminal chain of the phenylindole to the rings of Tyr408 and
His399. The 4-sulfamoyloxyphenyl moiety points to a
surface region of SpnHyal that is not occupied by Vcpal.
Inhibitory activity on Hyals
The inhibitory effects of Vc (1), Vcpal (2), and the
phenylindole inhibitor 4 were measured at pH 5.0 utilizing two Hyal
enzymes, SagHyal from strain 4755 and SpnHyal.
Concentrations for 50% inhibition (IC50) are summarzied in Table II.
Vc was only weakly active against both enzymes with IC50
values of 6 mM for SagHyal and 32 mM for SpnHyal under
our reaction conditions, respectively. On the contrary,
Vcpal inhibited SagHyal with an IC50 of 4 μM, and the
inhibition of SpnHyal was about 25-fold lower (Botzki
et al., 2004). The phenylindole compound, 4, inhibits
SagHyal with IC50 of 11 μM. Its IC50 could not be
determined against SpnHyal because the phenylindole has
limited solubility under our assay conditions. Following the
trend of SagHyal “selectivity” observed on Vc and Vcpal,
an IC50 value of compound 4 against SpnHyal was
estimated to be between 200 and 300 μM.
It is interesting to consider the possible reasons for the
markedly lower inhibition of SpnHyal compared with
SagHyal. The overall sequence identity of the crystallized
Hyals SpnHyal (Li et al., 2000) and SagHyal from strain
3502 (Li and Jedrzejas, 2001) is only 53%, but the amino
acids of the inhibitor-binding sites are essentially identical.
The structures of the α- and the following β-domains of
both enzymes are similar and have rms deviations of 1.54
and 1.18 Å, respectively. The flexibility behavior of both
enzymes was analyzed by molecular dynamics (Jedrzejas
et al., 2002, Mello et al., 2002), indicating similar major
modes of motion, (1) rotation/twisting of the whole
αdomain relative to the top half of the β-domain and (2)
opening/closing of the substrate-binding cleft (Jedrzejas
et al., 2002; Mello et al., 2002). Nevertheless, there may be
differences in the free energy of the conformational changes
leading to different substrate kinetics (apparent Km, Vmax)
and/or inhibitor affinity. However, because Vmax does not
differ significantly between the two enzymes and because a
reduction of the HA concentration did not significantly
lower IC50 values, the weaker SpnHyal activity of all
ND, not determined.
The assays were performed at pH 5.0 as described in the text.
aSagHyal from strain 4755.
bThe standard error of the mean IC50 (SEM) of four experiments was in
all cases lower than 10%.
cThe data of Vc and Vcpal are from Botzki and others (2004).
dNot soluble under assay conditions in concentrations >50–100 μM.
compounds should be rather attributed to the inhibitors
themselves, for example, to slower association or higher
dissociation rates because of the specific topology around
the binding cleft or the flexibility of the enzyme–inhibitor
complex. From Figure 5A, it becomes obvious that the
inhibitors occupy only about half of the
hexasaccharidebinding site, especially the product exit position where HA1
is located. The inner positive patch that interacts with HA3
disaccharide of the HA substrate (Figure 1) does not
contribute to inhibitor binding. Motion mode 1 (see above)
likely triggers the one-disaccharide unit shift of HA via
specific conformational changes of the positive patch relative
to residues from the active site and the negative patch
(Jedrzejas et al., 2002; Mello et al., 2002). Thus, the
flexibility of the enzyme–substrate and the enzyme–inhibitor
complexes may differ.
Structure-based design of benzoxazole-2-thiones
as Hyal inhibitors
The knowledge of the binding modes of Vcpal and the
phenylindole 4 within the active site of SpnHyal provides
several options for the design of more potent inhibitors.
The combination of a small hydrophilic molecule like Vc
or 6-sulfamoyloxyindole with a larger lipophilic residue
like the hexadecanoic or the decyl moiety appears to be an
appropriate strategy. Mainly hydrophilic, but more
potent, inhibitors than Vc can serve as optimal anchors in
the center of the binding site. Such compounds have been
suggested by structure-based ligand design methods
(Botzki et al., 2005). The most active inhibitor proposed
by the program LUDI (Böhm, 1992) was
1,3-diacetylbenzimidazole-2-thione (3, Figure 2). Replacement of one of
the N-acetyl fragments by oxygen has led to the
benzoxazole-2-thione scaffold, which may be suggested to
overlay with the indole moiety of compound 4. If this is the
case, alkanoyl chains at the nitrogen should point into the
same direction like the aliphatic tail of 4. Indeed,
benzoxazolethione derivatives 5 (Figure 2) with different
Nalkanoyl substituents were found to inhibit SagHyal from
strain 4755. Comparison of their putative binding mode
with that of the phenylindole 4, Vcpal, and the
hexasaccharide substrate was used to optimize the
Docking of the benzoxazole-2-thione scaffold into the
SpnHyal-phenylindole 4 complex in close superposition
with the indole moiety of compound 4 suggests additional
interactions (Figure 5C). The benzoxazole oxygen forms a
H bond with the amide NH2 group of Asn290, and the
thione sulfur is 2.65 Å distant from the indole NH of
Trp291. The oxygen of the N-alkanoyl substituents
interacts with the guanidine moiety of Arg462, with the OH
group of Tyr408 and possibly also with a protonated Nτ
imidazole atom of His399. If one projects the
benzoxazolethione scaffold in this position into the
SpnHyal-hexasaccharide complex, conclusions about optimal substituents can
be derived. As shown in Figure 5C, 3-phenyl substituted
Npropionyl derivatives are promising because the region at the
1,3-glycosidic linkage within HA1 corresponds to the two
methylene groups and because the N-acetylglucosamine
residue of HA1 overlaps with the terminal phenyl ring.
Results on SagHyal strain 4755 inhibition indeed show
that the 3-phenylpropionyl derivative is more potent than
its acetyl and butyryl homologs (Table II). The position of
the 3-phenylpropionyl moiety also corresponds to the
methylene groups 2–5 and 2–6 of the alkyl chains of the
phenylindole 4 and Vcpal, respectively. On the basis of
these comparisons, further derivatives with the common
scaffold 5 should share the following structural elements: a
benzoxazole moiety as anchor in the upper part of the
binding site, a propionic acid as spacer and a variety of
ring-based 3-substituents such as morpholine, piperazine,
cyclohexyl, phenyl, and 4-carboxyphenyl occupying the
space of the N-acetylglucosamine residue. The synthesis of
the representatives of the new compounds is currently
Our continuing search for new and novel inhibitors of
streptococcal Hyal recently led to the structure-based
design, synthesis, and analyses of Vcpal (compound 2),
1,3-diacetylbenzimidazole-2-thione (compound 3), and
the most recent
1-decyl-2-(4-sulfamoyloxyphenyl)-1Hindol-6-yl sulfamate (compound 4). The last compound, 4,
with an IC50 value of 11 μM against SagHyal from strain
4755 is one of the most potent inhibitors of this enzyme
known to date. The X-ray structure of the SpnHyal–
phenylindole 4 complex demonstrated this compound
binding in the active site of the enzyme and rationalized its
competitive nature. Interactions with catalytic Asn349
and the amino acid residues of the catalytically essential
hydrophobic patch were the basis for the inhibitory activity.
On the basis of the binding modes of the phenylindole 4,
a new compound,
N-(3-phenylpropionyl)-benzoxazole-2thione, was predicted and confirmed as yet another strong
Hyal inhibitor (IC50 value of 15 μM at SagHyal). All Hyal
inhibitors described are being further investigated for
their antibacterial properties in vivo. The inhibition of the
spread of pathogenic Streptococci, especially in the early
stage of their infections, will prevent the onset of bacterial
disease, namely bacteremia and meningitis. The
prospective importance of Hyal as target for the adjuvant therapy
of such diseases prompts our further search for new
inhibitors. In this respect, the binding modes of the
phenylindole and the benzoxazole-2-thione lead structures provide
the necessary information for further structure-based
inhibitor design to optimize potency, selectivity, and
Materials and methods
Hyaluronan HA from Streptococcus zooepidemicus was
purchased from Aqua Biochem (Dessau, Germany). Bovine
serum albumin (BSA) was obtained from Serva (Heidelberg,
Germany). All other chemicals were of analytical grade and
were obtained either from Merck (Darmstadt, Germany),
from Fisher Scientific (Pittsburgh, PA), or Sigma Chemical
(St. Louis, MO).
Enzymes and inhibitor preparations
Stabilized SagHyal, that is, 200,000 IU of the lyophilized
enzyme (0.572 mg from strain 4755, plus 2.2 mg BSA and
37 mg Tris–HCl per vial according to the information from
the supplier) was kindly provided by id-Pharma (Jena,
Germany). S. pneumoniae Hyal (Jedrzejas et al., 1998;
Jedrzejas, 2000, 2001; Li et al., 2000) was produced as
previously described (Jedrzejas, et al., 1998). The enzyme was
concentrated to 5 mg/mL in 10 mM Tris-HCl buffer, 150
mM NaCl, 1 mM dithiothreitol (DTT) (pH 7.4) using
centrifugal spin devices with a 50-kDa molecular weight
cutoff (Millipore, Billerica, MA). This preparation was used
for inhibition studies and for the production of crystals of
complex with inhibitor. Enzyme concentration was
determined photometrically at 280 nm using a calculated molar
absorption coefficient of 127,090 L mol−1 cm−1 (Jedrzejas et
al., 1998) according to Pace and others (1995). L-ascorbic
acid-6-hexadecanoate was obtained as described (Botzki,
et al., 2004).
1-decyl-2-(4-sulfamoyloxyphenyl)-1H-indol6-yl sulfamate (4) (Walter et al., 2004) and the
benzoxazole2-thiones 5a–5c were synthesized in our laboratory as
described (Braun, 2005).
Determination of inhibitory activity on Hyals
The inhibitory effects of the compounds on the Hyals from
S. pneumoniae and S. agalactiae strain 4755 were measured
using a turbidimetric assay (Di Ferrante, 1956). Enzyme
activity was quantified by determining the turbidity caused
by the residual high molecular weight substrate (molecular
weight >6–8 kDa) precipitated with cetyltrimethyl
ammonium bromide (CTAB). The incubation mixture contained
120 μL of citrate-phosphate buffer (solution A: 0.1 M
Na2HPO4/0.1 M NaCl, solution B: 0.1 M citric acid/0.1 M
NaCl, and solutions A and B were mixed in appropriate
proportions to reach pH 5.0), 30 μl of BSA solution (0.2
mg/mL in water), 30 μL of HA substrate solution (2 mg/mL
in water), 50 μL of H2O, 10 μL of dimethyl sulfoxide
(DMSO), and 30 μL of enzyme solution. The assays were
normalized by using equiactive concentrations of 2.9 ng of
Hyal from S. agalactiae strain 4755 or 1.7 ng of Hyal from
S. pneumoniae in 30 μL of BSA solution. To determine the
enzyme activities in the presence of the test compounds,
instead of 10 μL of DMSO, 10 μL of varying
concentrations of inhibitors were used. The final DMSO
concentration was 3.8% (v/v).
After incubation of the assay mixture for 30 min at 37°C,
720 μL of a 2.5% (m/v) cetyltrimethylammonium bromide
solution (2.5 g of CTAB dissolved in 100 mL of 0.5 M
sodium hydroxide solution, pH 12.5) was added to
precipitate the residual high molecular weight substrate and to
stop the enzyme reaction. This mixture was again incubated
at 25°C for 20 min, and the turbidity of each sample was
determined at 600 nm with an Uvikon 930 UV
spectrophotometer (Kontron, Eching, Germany). Experiments were
performed in quadruplicate. The turbidity of the sample
without inhibitor was taken as reference for 100% enzyme
activity, whereas the turbidity of the sample without
enzyme (30 μL BSA) was taken as reference for 0% enzyme
activity. The activities were plotted against the logarithm of
the inhibitor concentration, and IC50 values were calculated
by curve fitting of experimental data with Sigma Plot 8.0
(SPSS, Chicago, IL).
Molecular models of S. pneumoniae Hyal in complex
Crystallization of the complex
The hanging drop vapor diffusion method (McPherson, 1999)
using Linbro culture plates (Hampton Research, Aliso Viecho,
CA) at room temperature was used to grow the crystals of the
enzyme–inhibitor complex. Equal volumes of protein,
reservoir solution (1 μL each), and various amounts of inhibitor
solutions (0.1, 0.5, and 1.0 μL) were mixed and equilibrated
against 1 mL of the reservoir solution which contained 60–
65% saturated ammonium sulfate, 0.2 M NaCl, 2% dioxane,
and 0.1 M sodium citrate buffer (pH 6.0) as reported for the
native enzyme crystallization (Jedrzejas et al., 1998).
The crystals of the enzyme–inhibitor complex were
cryoprotected using 30% xylitol (w/v), 3.5 M ammonium
sulfate, and 0.1 M sodium citrate buffer (pH 6.0) as reported
for the native crystals (Li et al., 2000) and frozen in liquid
nitrogen. Standard fiber loops (Hampton Research) of a
suitable size were used to pick up and mount the frozen
crystals under nitrogen flow at –180°C. The X-ray
diffraction for the inhibitor complex data was collected using
rotation (oscillation) photography and a Quantum 4u CCD
detector. The crystallographic setup of beamline 5.0.1 of
the Berkeley Center for Structural Biology, Advanced
Light Source, Lawrence Berkeley National Laboratory was
used. The collected data were analyzed, indexed, integrated,
and scaled using the HKL2000 software package (Otwinowski
and Minor, 1997). The resultant crystals were isomorphous
to the native S. pneumoniae Hyal crystals (Jedrzejas, 1998).
Structure solution and refinement
The structure was solved by rigid body refinement in
Crystallography and NMR System (CNS) (Brunger et al., 1998)
using the crystal structure of S. pneumoniae Hyal in
complex with Vcpal (Botzki et al., 2004) as a search model. The
iterative approach proceeded with alternating cycles of
computational refinement with CNS (Brunger et al., 1998)
and manual rebuilding using O (Jones, et al., 1991). No
intensity or sigma-based cut-offs were applied to the data.
SigmaA-weighted map coefficients (Read, 1986) were used
throughout. The Rfree value (Brunger, 1992), calculated for
a test set of 5% of reflections, was used to monitor the
progress of refinement. Difference density at the active site
was apparent even in initial maps. The aliphatic portion of
the inhibitor was relatively easily placed, but different
orientations of the remainder of the inhibitor were required.
The possibilities were assessed using the quality of electron
density maps, B-factors, and the Rfree value as criteria.
Sulfate and xylitol molecules, deriving from the
crystallization and cryo-cooling solutions, respectively, were modeled
into suitably shaped regions of electron density. Final
statistics for the model are summarized in Table I. Programs
of the CCP4 package (Collaborative Computational
Project, Number 4, 1994) were used for manipulations and
structural superpositions made with LSQMAN (Kleywegt,
1999). PYMOL was used to generate illustrations (DeLano,
Modeling approaches were performed using the
BIOPOLYMER module of the software package SYBYL 6.9 (Tripos,
St. Louis, MO). Energy minimizations were based on the
AMBER all-atom force field (Cornell et al., 1995) with
AMBER all-atom charges (distance-dependent dielectric
constant of 4) up to an rms gradient of 0.1 kcal/mole/Å (Powell
conjugate gradient). Surface areas and lipophilic potentials (protein
variant with the new Crippen parameter table [Heiden et al.,
1993; Ghose et al., 1998]) of the model were calculated and
visualized by the program MOLCAD (Brickmann et al., Tech.
University Darmstadt, Germany) contained within SYBYL.
The diffraction data were collected at the Berkeley Center
for Structural Biology, Advanced Light Source, Lawrence
Berkeley National Laboratory using beamline 5.0.1. The
coordinates and structural factors of the SpnHyal–inhibitor
complex have been deposited to the PDB with accession
numbers 2BRP and PDB2BRP for the coordinates and
structure factors, respectively. This study was supported in
part by NIH grant AI44079 (M.J.J.), by the Graduate
College 760 of the Deutsche Forschungsgemeinschaft, and by
the Studienstiftung des Deutschen Volkes (A.B.).
Conflict of interest statement
BSA, bovine serum albumin; CTAB, cetyltrimethyl
ammonium bromide; DMSO, dimethyl sulfoxide; DTT,
dithiothreitol; ECM, extracellular matrix; HA, hyaluronan; Hyal,
hyaluronan lyases; IC50, concentration for 50% inhibition;
PDB, Protein Data Bank; rms, root mean square; SagHyal,
Streptococcus agalactiae hyaluronan lyase; SpnHyal,
Streptococcus pneumoniae hyaluronan lyase; Vc, L-ascorbic acid;
Vcpal, L-ascorbic acid-6-hexadecanoate.
Berman , H.M. , Westbrook , J. , Feng , Z. , Gilliland , G. , Bhat , T.N. , Weissig , H. , Shindyalov , I.N. , and Bourne , P.E. ( 2000 ) The Protein Data Bank . Nucleic Acids Res ., 28 , 235 - 242 .
Böhm , H.J. ( 1992 ) The computer program LUDI: a new method for the de novo design of enzyme inhibitors . J. Comput. Aided Mol. Des ., 6 , 61 - 78 .
Botzki , A. , Rigden , D.J. , Braun , S. , Nukui , M. , Salmen , S. , Hoechstetter , J. , Bernhardt , G. , Dove , S. , Jedrzejas , M.J. , and Buschauer , A. ( 2004 ) L-Ascorbic acid, 6-hexadecanoate, a potent hyaluronidase inhibitor. X-ray structure and molecular modeling of enzyme-inhibitor complexes . J. Biol. Chem. , 279 , 45990 - 45997 .
Botzki , A. , Salmen , S. , Bernhardt , G. , Buschauer , A. , and Dove , S. ( 2005 ) Structure-based design of bacterial hyaluronan lyase inhibitors . QSAR Comb. Sci. , 24 , 458 - 469 .
Braun , S. ( 2005 ) New Inhibitors of Bacterial Hyaluronidases - Synthesis and Structure-Activity Relationships . University of Regensburg Press, Regensburg, Germany.
Brunger , A.T. ( 1992 ) The free R value: a novel statistical quantity for assessing the accuracy of crystal structures . Nature , 355 , 472 - 474 .
Brunger , A.T. , Adams , P.D. , Clore , G.M. , DeLano , W.L. , Gros , P. , Grosse-Kunstleve , R.W. , Jiang , J.S. , Kuszewski , J. , Nilges , M. , and others ( 1998 ) Crystallography & NMR system: a new software suite for macromolecular structure determination . Acta Crystallogr. D Biol . Crystallogr., 54 , 905 - 921 .
Collaborative Computational Project , Number 4. ( 1994 ) The CCP4 suite: programs for protein crystallography . Acta Crystallogr. D Biol . Crystallogr., 50 , 760 - 763 .
Cornell , W.D. , Cieplak , P. , Bayly , C.I. , Gould , I.R. , Merz , K.M. Jr., Ferguson , D.M. , Spellmeyer , D.C. , Fox , T. , Caldwell , J.W. , and Kollman , P.A. ( 1995 ) A second generation force field for the simulation of proteins, nucleic acids, and organic molecules . J. Am. Chem. Soc. , 117 , 5179 - 5197 .
Coutinho , P.M. and Henrissat , B. ( 1999 ) Carbohydrate-Active Enzymes server Available at: http://afmb.cnrs-mrs. fr/CAZY/ Accessed March 2006 .
DeLano , W.L. ( 2003 ) The PyMOL Molecular Graphics System . www.pymol.org, San Carlos, USA Accessed March, 2006 .
Di Ferrante , N. ( 1956 ) Turbidimetric measurement of acid mucopolysaccharides and hyaluronidase activity . J. Biol. Chem . 220 , 303 - 306 .
Dillon , H.C. Jr., Khare , S. , and Gray , B.M. ( 1987 ) Group B streptococcal carriage and disease: a, 6-year prospective study . J. Pediatr 110 , 31 - 36 .
Gallivan , J.P. and Dougherty , D.A. ( 1999 ) Cation-pi interactions in structural biology . Proc. Natl. Acad. Sci. U. S. A. , 96 , 9459 - 9464 .
Ghose , A.K. , Viswanadhan , V.N. , and Wendoloski , J.J. ( 1998 ) Prediction of hydrophobic (lipophilic) properties of small organic molecules using fragmental methods: An analysis of ALOGP and CLOGP methods . J. Phys. Chem. , 102 , 3762 - 3772 .
Heiden , W. , Moeckel , G. , and Brickmann , J. ( 1993 ) A new approach to analysis and display of local lipophilicity/hydrophilicity mapped on molecular surfaces . J. Comput. Aided Mol. Des ., 7 , 503 - 514 .
Hynes , W. and Walton , S. ( 2000 ) Hyaluronidases of Gram-positive bacteria . FEMS Microbiol. Lett. , 183 , 201 - 207 .
Jedrzejas , M.J. ( 2000 ) Structural and functional comparison of polysaccharidedegrading enzymes . Crit. Rev. Biochem. Mol. Biol ., 35 , 221 - 251 .
Jedrzejas , M.J. ( 2001 ) Pneumococcal virulence factors: structure and function . Microbiol. Mol Biol. Rev. , 65 , 187 - 207 .
Jedrzejas , M.J. ( 2002a ) Mechanisms of polysaccharide degradation by bacterial enzymes: Degradation of Hyaluronan . In Pandali, S.G. (ed.), Recent Research Developments in Biophysics and Biochemistry. Research Signpost, Trivandrum, India , pp. 197 - 225 .
Jedrzejas , M.J. ( 2002b ) Three-dimensional structures of hyaluronate lyases from Streptococcus species and their mechanism of hyaluronan degradation . In Hascall, V.C. and Yanagishita , M. (eds), Science of hyaluronan today . Glycoforum Available at: www.glycoforum.gr.jp/ science/hyaluronan. Accessed March , 2006 .
Jedrzejas , M.J. and Chantalat , L. ( 2000 ) Structural studies of Streptococcus agalactiae hyaluronate lyase . Acta Crystallogr. D Biol . Crystallogr., 56 , 460 - 463 .
Jedrzejas , M.J. , Chantalat , L. , and Mewbourne , R.B. ( 1998 ) Crystallization and preliminary X-ray analysis of Streptococcus pneumoniae hyaluronate lyase . J. Struct. Biol ., 121 , 73 - 75 .
Jedrzejas , M.J. , Mello , L.V. , De Groot , B.L. , and Li , S. ( 2002 ) Mechanism of hyaluronan degradation by Streptococcus pneumoniae hyaluronate lyase: Structures of complexes with the substrate . J. Biol. Chem. , 277 , 28287 - 28297 .
Jedrzejas , M.J. , Mewbourne , R.B. , Chantalat , L. , and McPherson , D.T. ( 1998 ) Expression and purification of Streptococcus pneumoniae hyaluronate lyase from Escherichia coli . Protein Expr. Purif. , 13 , 83 - 89 .
Jones , T.A. , Zou , J.Y. , Cowan , S.W. , and Kjeldgaard , M. ( 1991 ) Improved methods for building protein models in electron density maps and the location of errors in these models . Acta Crystallogr. A , 47 , 110 - 119 .
Kelly , S.J. , Taylor , K.B. , Li , S. , and Jedrzejas , M.J. ( 2001 ) Kinetic properties of Streptococcus pneumoniae hyaluronate lyase . Glycobiol., 11 , 297 - 304 .
Klein , E. and Weber , N. ( 2001 ) In vitro test for the effectiveness of antioxidants as inhibitors of thiyl radical-induced reactions with unsaturated fatty acids . J. Agric. Food Chem ., 49 , 1224 - 1227 .
Kleywegt , G.J. ( 1999 ) Experimental assessment of differences between related protein crystal structures . Acta Crystallogr. D Biol . Crystallogr., 55 , 1878 - 1884 .
Kreil , G. ( 1995 ) Hyaluronidases-a group of neglected enzymes . Protein Sci. , 4 , 1666 - 1669 .
Laurent , T. and Fraser , J. ( 1992 ) Hyaluronan . FASEB J. , 6 , 2397 - 2405 .
Li , S. and Jedrzejas , M.J. ( 2001 ) Hyaluronan binding and degradation by Streptococcus agalactiae hyaluronate lyase . J. Biol. Chem. , 276 , 41407 - 41416 .
Li , S. , Kelly , S. , Lamani , E. , Ferraroni , M. , and Jedrzejas , M.J. ( 2000 ) Structural basis of hyaluronan degradation by Streptococcus pneumoniae hyaluronate lyase . EMBO J ., 19 , 1228 - 1240 .
Li , S. , Taylor , K.B. , Kelly , S.J. , and Jedrzejas , M.J. ( 2001 ) Vitamin C inhibits the enzymatic activity of Streptococcus pneumoniae hyaluronate lyase . J. Biol. Chem. , 276 , 15125 - 15130 .
Linker , A. and Meyer , K. ( 1954 ) Production of unsaturated uronides by bacterial hyaluronidases . Nature , 174 , 1192 - 1194 .
McPherson , A. ( 1999 ) Crystallization of Biological Molecules . Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.
Mello , L.V. , De Groot , B.L. , Li , S. , and Jedrzejas , M.J. ( 2002 ) Structure and flexibility of Streptococcus agalactiae hyaluronate lyase complex with its substrate. Insights into the mechanism of processive degradation of hyaluronan . J. Biol. Chem. , 277 , 36678 - 36688 .
Meyer , K. and Palmer , J.W. ( 1934 ) The polysaccharide of the vitreous humor . J. Biol. Chem. , 107 , 629 - 634 .
Mitra , A. , Govindwar , S. , Joseph , P. , and Kulkarni , A. ( 1992 ) Inhibition of human term placental and fetal liver glutathione-S-transferases by fatty acids and fatty acid esters . Toxicol. Lett. , 60 , 281 - 288 .
Nukui , M. , Taylor , K.B. , McPherson , D.T. , Shigenaga , M.K. , and Jedrzejas , M.J. ( 2003 ) The function of hydrophobic residues in the catalytic cleft of Streptococcus pneumoniae hyaluronate lyase . Kinetic characterization of mutant enzyme forms . J. Biol. Chem. , 278 , 3079 - 3088 .
Otwinowski , Z. and Minor , W. ( 1997 ) Processing of X-ray diffraction data collected in oscillation mode . Methods Enzymol. , 276 , 307 - 326 .
Pace , C.N. , Vajdos , F. , Fee , L. , Grimsley , G. , and Gray , T. ( 1995 ) How to measure and predict the molar absorption coefficient of a protein . Protein Sci. , 4 , 2411 - 2423 .
Ponnuraj , K. and Jedrzejas , M.J. ( 2000 ) Mechanism of hyaluronan binding and degradation: structure of Streptococcus pneumoniae haluronate lyase in complex with hyaluronic acid ddisaccharide at 1.7 Å resolution . J. Mol. Biol ., 299 , 885 - 895 .
Read , R.J. ( 1986 ) Improved Fourier coefficients for maps using phases from partial structures with errors . Acta Crystallogr. A. , 42 , 140 - 149 .
Rigden , D.J. and Jedrzejas , M.J. ( 2003 ) Genome-based identification of a carbohydrate binding module in Streptococcus pneumoniae hyaluronate lyase . Proteins , 52 , 203 - 211 .
Rigden , D.J. , Galperin , M.Y. , and Jedrzejas , M.J. ( 2003 ) Analysis of structure and function of putative surface-exposed proteins encoded in the Streptococcus pneumoniae genome: a bioinformatics-based approach to vaccine and drug design . Crit. Rev. Biochem. Mol. Biol ., 38 , 143 - 168 .
Salmen , S. ( 2003 ) Inhibitors of Bacterial and Mammalian Hyaluronidases: Synthesis and Structure-Activity Relationships . University of Regensburg Press, Regensburg, Germany.
Simpson , G.L. and Ortwerth , B.J. ( 2000 ) The non-oxidative degradation of ascorbic acid at physiological conditions . Biochim. Biophys. Acta , 1501 , 12 - 24 .
Suzuki , K. , Terasaki , Y. , and Uyeda , M. ( 2002 ) Inhibition of hyaluronidases and chondroitinases by fatty acids . J. Enzyme Inhib Med. Chem. , 17 , 183 - 186 .
Walter , G. , Liebl , R., and von Angerer , E. ( 2004 ), 2 - phenylindole sulfamates: inhibitors of steroid sulfatase with antiproliferative activity in MCF-7 breast cancer cells . J. Steroid Biochem. Mol. Biol ., 88 , 409 - 420 .