Heterocyclic Bis-Cations as Starting Hits for Design of Inhibitors of the Bifunctional Enzyme Histidine-Containing Protein Kinase/Phosphatase from Bacillus subtilis

Article abstract of DOI:10.1021/jm021043o

The main mechanism of carbon catabolite repression/activation in low-guanine and low-cytosine Gram-positive bacteria seems to involve phosphorylation of HPr (histidine-containing protein) at Ser-46 by the ATP-dependent HPr kinase, which in Bacillus subtilis, Lactobacillus casei, and Staphylococcus xylosus also exhibits phosphatase activity and is thus a bifunctional enzyme (HPrK/P). Since deficiency of HPrK/P in S. xylosus, L. casei, and B. subtilis mutants leads to severe growth defects, inhibitors of the enzyme could form a new family of antibiotic drugs. The aim of the study was to screen an in-house chemical library for identification of hits as inhibitors of HPrK/P in B. subtilis and to further extract additional information of structural features from hit optimization using a radioactive in vitro assay. A symmetrical bis-cationic compound LPS 02-10-L-D09 (2a) with a 12-carbon alkyl linker bridging the two 2-aminobenzimidazole moieties was identified as a non-ATP mimetic compound exhibiting an EC₅₀ value of 10 μM in a kinase assay with HPr as substrate. The substance also inhibited the phosphatase activity of HPrK/P triggered by the addition of inorganic phosphate. Similar results were obtained with 2a and catabolite repression HPr, which, like HPr, can be phosphorylated at Ser-46 by HPrK/P and is involved in catabolite repression. Structure-activity relationship analysis indicated the importance in its structure of a substituted 2-aminobenzimidazole. This typical heterocycle is linked through a C12 alkyl chain to a second scaffold that can bear a cationic or a noncationic moiety but in all cases should present an aromatic ring in its vicinity.

Full text of DOI:10.1021/jm021043o

2264
J . Med. Chem. 2004, 47, 2264-2275
Heter ocyclic Bis-Ca tion s a s Sta r tin g Hits for Design of In h ibitor s of th e
Bifu n ction a l En zym e Histid in e-Con ta in in g P r otein Kin a se/P h osp h a ta se fr om
Ba cillu s su btilis
Helena Ramstro¨m,§,† Maryline Bourotte,§,‡ Claude Philippe,† Martine Schmitt,‡ J acques Haiech,† and
J ean-J acques Bourguignon*,‡
Pharmacologie et Physico-Chimie des Interactions Cellulaires et Mole´culaires and Laboratoire de Pharmacochimie de la
Communication Cellulaire, Universite´ Louis Pasteur de Strasbourg, Faculte´ de Pharmacie, 74 Route du Rhin, B.P. 24,
F-67401 Illkirch, France
Received September 12, 2002
The main mechanism of carbon catabolite repression/activation in low-guanine and low-cytosine
Gram-positive bacteria seems to involve phosphorylation of HPr (histidine-containing protein)
at Ser-46 by the ATP-dependent HPr kinase, which in Bacillus subtilis, Lactobacillus casei,
and Staphylococcus xylosus also exhibits phosphatase activity and is thus a bifunctional enzyme
(HPrK/P). Since deficiency of HPrK/P in S. xylosus, L. casei, and B. subtilis mutants leads to
severe growth defects, inhibitors of the enzyme could form a new family of antibiotic drugs.
The aim of the study was to screen an in-house chemical library for identification of hits as
inhibitors of HPrK/P in B. subtilis and to further extract additional information of structural
features from hit optimization using a radioactive in vitro assay. A symmetrical bis-cationic
compound LPS 02-10-L-D09 (2a ) with a 12-carbon alkyl linker bridging the two 2-aminoben-
zimidazole moieties was identified as a non-ATP mimetic compound exhibiting an EC50 value
of 10 µM in a kinase assay with HPr as substrate. The substance also inhibited the phosphatase
activity of HPrK/P triggered by the addition of inorganic phosphate. Similar results were
obtained with 2a and catabolite repression HPr, which, like HPr, can be phosphorylated at
Ser-46 by HPrK/P and is involved in catabolite repression. Structure-activity relationship
analysis indicated the importance in its structure of a substituted 2-aminobenzimidazole. This
typical heterocycle is linked through a C12 alkyl chain to a second scaffold that can bear a
cationic or a noncationic moiety but in all cases should present an aromatic ring in its vicinity.
In tr od u ction
comitant transport of PTS carbohydrates.16 When HPr
is phosphorylated by enzyme I (EI) at His-15, the
phosphoryl group can then be transferred to the sugar-
specific enzymes II (EIIAs) of the PTS system.16 HPr of
low-GC Gram-positive bacteria can also be phosphory-
lated at Ser-46 in the presence of ATP by HPr kinase
in a mechanism that exclusively serves regulatory
purposes.7-11 In addition to its kinase activity, it has
been shown for Enterococcus faecalis,12 B. subtilis,12
Lactobacillus casei,13 and Staphylococcus xylosus14 that
the enzyme also possesses HPr(Ser-P) phosphatase
activity, which becomes prevalent in the presence of
inorganic phosphate. Thus, in these species it is a
bifunctional enzyme (HPrK/P). Furthermore, HPrK/P
of B. subtilis has been found to be a homo-oligomeric
enzyme that shows strong positive cooperativity for
fructose-1,6-bisphosphate (FBP), ATP, and other nucle-
otides.17,18
The major intracellular regulatory mechanism in
many biological processes involves phosphorylation and
dephosphorylation of proteins by protein kinases and
protein phosphatases.1-3 Since the discovery of protein
phosphorylation in the mid-1950s in the case of rabbit
skeletal muscle glycogen phosphorylase,4 a large num-
ber of protein kinases and phosphatases have been
characterized in a variety of systems.1,2 For prokaryotes,
the first conclusive evidence for the existence of protein
kinase activity was presented in the late 1970s in
Salmonella typhimurium5 but is now well-known in a
number of other bacterial species.6
In low-GC (guanine, cytosine) Gram-positive bacteria
such as Bacillus subtilis, the main regulatory mecha-
nism in the hierarchical use of carbon sources seems to
involve phosphorylation of HPr (histidine-containing
protein) at Ser-46 by the ATP-dependent HPr kinase.7-15
HPr is a small phosphocarrier protein of the bacterial
phosphoenolpyruvate (PEP)/sugar phosphotransferase
system (PTS) that catalyzes phosphorylation and con-
In a complex medium with different carbohydrates,
the bacteria preferentially utilizes the one that allows
the fastest growth, e.g., glucose, whereby the expression
of secondary catabolic genes is repressed. The regulatory
mechanism involves activation of the kinase of HPrK/
P, which is typically stimulated by glycolytic intermedi-
ates such as FBP, although not in E. faecalis12 and
Streptococcus salivarius,15 with concomitant phospho-
rylation of HPr at Ser-46.7-15 HPr(Ser-P) then forms a
* To whom correspondence should be addressed. Phone: +33 (0)3
90 24 42 34. Fax: +33 (0)3 90 24 43 10. E-mail: jjb@aspirine.u-
strasbg.fr.
§ The first two authors contributed equally to this work.
† Pharmacologie et Physico-Chimie des Interactions Cellulaires et
Mole´culaires.
‡ Laboratoire de Pharmacochimie de la Communication Cellulaire.
10.1021/jm021043o CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/26/2004

Inhibitors of HPr Kinase/ Phosphatase in B. subtilis
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9 2265
complex with CcpA (catabolite control protein A),19
a
member of the GalR/LacI family of transcriptional
repressors/activators,20 and this complex interacts with
cis-active operator sequences, cres (catabolite responsive
elements).21,22 Carbon catabolite repression (CCR)23-26
and also carbon catabolite activation (CCA) of, for
example, glycolysis and carbon overflow metabolism27-29
are assumed to be the consequence of this complex
binding to DNA.
Another protein, Crh (catabolite repression HPr), for
which the gene was discovered within the B. subtilis
genome sequencing program,30 exhibits 45% sequence
identity with HPr.31 Crh can also be phosphorylated at
Ser-46 by HPrK/P and has been shown to be implicated
in CCR.24,31,32 However, His-15 is replaced by a Gln
residue that prevents phosphorylation by PEP and EI
and is therefore probably not involved in the PTS
transport of sugar.31 Thus, Crh is presumed to carry out
exclusively regulatory functions.31
Since HPrK/P-deficient S. xylosus,14 L. casei,13 and B.
subtilis HPrK/P mutants33 as well as mutants affecting
in particular the phosphatase activity34,35 lead to severe
growth defects and since the main mechanism of CCR/
CCA is suggested to involve the enzyme HPrK/P,
inhibitors of HPrK/P could form a new family of
antibiotic drugs. The need for new antibiotics can be
seen in light of the increasing problem of pathogenic
bacteria resistant to most known antibiotic drugs. For
example, more than 95% of Staphylococcus aureus
strains worldwide are resistant to penicillin and up to
60% are resistant to methicillin.36 Although efforts are
made to reduce the spread of antibiotic resistance, the
problem still remains. The current situation strengthens
the need for novel types of antibacterial agents with
unique targets such that cross-resistance does not
readily occur.
F igu r e 1. (A) Chemical structure of LPS 02-10-L-D09 (2a )
(oxalate salt). (B) Phosphorylation assay with HPr (10 µM)
after adding different amounts of 2a . Data represent the mean
( SEM (n ) 3). The EC50 value was determined to be 10 µM.
(C) Dephosphorylation assay with HPr(Ser-P) (8 µM). After
complete phosphorylation, inorganic phosphate (K2HPO4 and
KH2PO4) was added to give a final concentration of 1 mM
together with different concentrations of 2a . The mixture was
further incubated at 37 °C for 2 h. Addition of inorganic
phosphate together with the three highest concentrations of
2a to the assay mixtures led to precipitation, and these
samples were excluded. The EC50 value was determined to be
14 µM. Data represent the mean ( SEM (n ) 3).
The rationale for building a focused library deals with
small peptides able to mimic the residue Ser-46, taking
into consideration both surrounding amino acids (Lys-
Ser-Ile). However, the most potent compound exhibited
moderate inhibitory capacity (EC50 ≈ 500 µM). Recently,
a chemical library of synthetic compounds (see Experi-
mental Section) originated from different teams and
medicinal chemistry projects were established within
the Laboratoire de Pharmacochimie de la Communica-
tion Cellulaire CNRS UMR 7081. The substances con-
stituting the library were designed in the past as
putative ligands of various central nervous system
(CNS) and non-CNS targets including enzymes, recep-
tors, and transport systems. Thus, a large chemical
diversity (various heterocyclic compounds, synthetic
amino acid derivatives, small peptides, etc.) could be
found within the in-house library. In addition to storage
in standard vials, the library is available for automated
screening stored in 96-well plates with 80 substances
per plate, allowing control samples in a total of 16 wells
(first and last columns). The initial dilution was made
with dimethyl sulfoxide (DMSO) to a final concentration
of around 30 mM, and the plates were stored under
argon at 4 °C. The structures of the compounds as well
as characteristics and other data are stored in a
relational database using the ISISBase software. The
first version (about 1500 compounds) was screened in
order to identify potential hits acting as HPrK/P inhibi-
Sch em e 1a
a
Reagents and conditions: (i) 0.5 equiv of Br(CH2)nBr.
tors. Among them, an interesting compound was identi-
) 10 µM) and served as
fied (2a in Figure 1A, EC50
starting hit for further structural optimization. The first
structure-activity relationship analysis is presented in
this paper.
Ch em istr y
The symmetrical bis-cationic reference compound 2a
and other related compounds were prepared as de-
scribed in Scheme 1. Depending on the nucleophilicity
of the starting heterocyclic amidine 1, the reaction was
performed in refluxing butanone, acetone, or sulfolane
in the presence of dibromoalkanes, affording the awaited

2266 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9
Ramstro¨m et al.
Sch em e 2a
value and the half and the double of this concentration.
Initial velocities (pmol/min) were plotted against HPr/
Crh concentrations.
In ter d om a in In ter a ction . The inhibitor was tested
at three different constant concentrations: the concen-
tration equal to its EC50 value, a lower concentration,
which caused only a small inhibition, and a concentra-
tion that gave maximal inhibition at different concen-
trations of ATP ranging from 1 µM to 1 mM. A change
in the Hill coefficient for ATP was used as an indication
of an interaction between the subunits.
Mu ltip le Dr u g Effect/Com bin a tion In d ex. For
multiple drug effects, the combination index (CI) method
described by Chou-Talalay was used.37 Briefly, two
inhibitors were included in the assay at equipotent
doses; i.e., the concentrations of the individual sub-
stances were varied, but the ratio between them was
kept constant, equal to the ratio at their EC50 values.
At least 13 different conditions, tested in triplicate, were
included. The CI was calculated at EC50 according to
the equation
a
Reagents and conditions: i: R-X, KOH, ethanol; ii: 2-bu-
tanone, 3 equiv. of Br(CH2)nR′ 7, 80°C, 48 h, 55-95%.
bis-cations 2a -h as dibromides in good yields (60-95%).
When the crude salt was not pure enough, the free base
could be chromatographed on silica gel and crystallized
as an oxalate (2a -d ). Because of the low reactivity for
N-substituted 2-aminobenzimidazole 4 toward dibro-
moalkanes, the reaction needed to be performed in
sulfolane in order to avoid significant amounts of the
monosubstituted derivative. The progressive alkylation
of the unsubsituted 2-aminobenzimidazole 4 using dif-
ferent functionalized halides R′(CH2)nX′ 7 was carried
out as described in Scheme 2. The alcohol derivatives
6r and 6s (R′ ) OH) served as intermediates because
they were submitted to oxidation in the presence of
chromium trioxide to afford the corresponding carboxylic
acids 8a and 8b. Removal of the N-Boc protection in
6w -y afforded the free amino derivatives 9a -c. Fi-
nally, the treatment of the bromo derivatives with
secondary amines yielded the unsymmetrical bis-cations
10a -i (Scheme 3).
D1
D2
RD1D2
CI )
+
+
Dx,1 Dx,2 Dx,1Dx,2
where D1 and D2 refer to the concentrations of each
individual substance when mixed together. The terms
Dx,1 and Dx,2 are the concentrations of the two different
substances when each compound is tested alone. The
value of R is 0 if the two compounds are mutually
exclusive and is 1 if they are mutually nonexclusive. The
tool to classify the effects of two compounds is the
median-effect plot, where log(dose) was plotted against
log(fraction phosphorylated/fraction nonphosphorylat-
ed). If the median-effect plots of each individual com-
pound, and when mixed together, are parallel or if the
compounds have similar modes of action, then R is 0.
However, if the plots of the individual compounds are
still parallel but the plot of the mixture of the two
compounds is upwardly concave, or if the two com-
pounds have independent modes of action, then R is 1.
A CI value less than 1 indicates synergism, a value
equal to 1 indicates summation, and a value greater
than 1 indicates antagonism.
En zym a tic Kin a se Assa y
HPr(His)6, Crh(His)6, and HPrK/P(Trx-His6-S-tag)
from B. subtilis were purified by Ni-NTA affinity
chromatography as described previously.17 The degree
of phosphorylation of HPr(His)6/Crh(His)6 (10 µM) by
HPrK/P(Trx-His6-S-tag) (100 nM) with and without the
addition of test compound was determined in an in vitro
kinase assay at pH 8.0 using ATP as the phosphoryl
donor and with the addition of 2 mM FBP and 0.1%
bovine serum albumine (BSA) (see the experimental
part).
EC50 Cu r ve. The EC50 (50% inhibition of phospho-
rylation/dephosphorylation) curves were based on 13
consecutively diluted concentrations of tested inhibitor
included in the assay before initiating the reaction with
the addition of enzyme. Typically, each concentration
was tested in triplicate. The volume was 2 µL of the
inhibitor dissolved in DMSO (10% v/v final DMSO
concentration). Control reaction mixtures contained 10%
v/v DMSO instead of the test compound.
ATP Com p etition Assa y. Different concentrations
of ATP (0.5, 0.75, 1, 2, and 3 mM) were included in the
assay with the addition of tested inhibitor at a constant
concentration equal to its EC50 value. This 6-fold
increase in ATP concentration would correspond to a
decrease in the inhibitor concentration to a similar
extent if the inhibitor is directly competing with ATP.
Since the nonspecific binding of ATP increased with
increasing concentration of ATP, the values for the
samples containing no enzyme (blank) were subtracted
from the corresponding values for the samples contain-
ing enzyme.
Resu lts
P h osp h or yla tion Assa y. To study the phosphory-
lation reaction of HPr(His6)/Crh(His6), a radioactive
assay for HPrK/P(Trx-His6-S-tag) was developed. The
phosphoryl donor [γ-32P]ATP was used together with an
excess of about 5 mM MgCl2. The final concentration of
FBP was 2 mM, since higher concentrations did not
increase the degree of phosphorylation. The reaction
was performed at pH 8 using Tris as the buffer and
initiated by a 100 times lower concentration of the
enzyme compared to protein substrates. Except for the
necessary ingredients for the reaction, BSA was in-
cluded in the assay to avoid nonspecific binding. A final
concentration of 0.1% BSA was used, and an increase
in concentration did not result in further improvement.
Parvalbumin was also evaluated but was less efficient
than BSA. After incubation, it was furthermore neces-
sary to immediately place the P81 papers containing the
samples in phosphoric acid solution. If the papers were
Distin ction of In h ibitor Typ e. HPr/Crh was used
over a concentration range from 0.1 to 200 µM with and
without three different fixed concentrations of tested
inhibitor, which are the concentration equal to its EC50

Inhibitors of HPr Kinase/ Phosphatase in B. subtilis
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9 2267
Sch em e 3a
a
Reagents and conditions: (i) CrO3, H2SO4, acetone/water, 30 min, 60-75%; (ii) HClg, AcOEt; (iii) acetonitrile, 3 equiv of HNR1R2, 85
°C, 48 h, 40-70%.
Ta ble 1. Characteristics of 2a Alone, ADP Alone, and When Mixed Togethera
EC50 (µM)
mean (95% confidence interval)
Hill slope
mean (95% confidence interval)
CI (R ) 0)
inhibitor
2a
ADP
2a + ADP
HPr
Crh
Kinase Assay
HPr
Crh
HPr
Crh
10 (8.6-12)
21 (17-25)
1.8 (1.3-2.2)
0.90 (0.5-1.3)
2.2 (1.9-2.5)
2.4 (1.5-3.3)
1.4 (1.2-1.7)
2.3 (1.9-2.7)
4300 (2400-7800)
5.2 (4.8-5.6) + 2100 (1900-2200)
7400 (6400-8700)
12 (11-13) + 4200 (3800-4600)
1.0
1.1
Phosphatase Assay
14 (8.5 to 23)
2a
16 (14-19)
2.1 (0.8-3.5)
1.7 (0.6-2.8)
a
Mean ()3) and 95% confidence intervals for the EC50 values and the Hill slope values were estimated using the program GraphPad
Prism. The following equations were used: Y ) Bottom + (Top - Bottom)/{1 + 10[X-log(EC )]Hill slope} for phosphorylation of HPr/Crh; Y )
50
Bottom + (Top - Bottom)/{1 + 10[log(EC )-X]Hill slope} for dephosphorylation of HPr/Crh. X is the logarithm of the concentration of inhibitor,
50
and Y is the response.
allowed to dry at room temperature before being washed
with phosphoric acid, a lower degree of phosphorylation
was obtained possibly because of the phosphate content
of the papers, which triggered the enzyme to act as a
phosphatase. Since the samples were prepared consecu-
tively, the time spent in the first wash with phosphoric
acid for the P81 papers was investigated. A longer time
spent in the first wash, up to 1 h, did not influence the
results significantly.
To obtain a fast, simple, and convenient screening
procedure, a somewhat modified kinase assay was used
in an automated 96-well format. Because the substances
in the library were diluted with DMSO, this solvent
needed to be included in the assay. Control experiment
revealed that 10% v/v of DMSO did not affect the degree
of phosphorylation. With use of a dispensing robot, the
phosphorylation reaction was terminated by adding 75
mM H3PO4 to the samples in the wells in the same order
as the reactions were initiated before spotting the
samples on P81 paper 96-well filter plates. This proce-
dure was performed to ensure that all enzymatic activity
was prevented on the P81 papers because the papers
contain phosphate and the robot worked column-by-
column. The washing procedure to remove unreacted
ATP from the P81 papers was subsequently carried out
with the plates mounted on a vacuum manifold.
P a r en t In h ibitor . The screening campaign was
performed with the different compounds from the in-
house library at a final concentration of approximately
30 µM. From this first screening, a single substance,
compound 2a (Figure 1A), was identified and submitted
to further investigations. It is a symmetrical bis-cationic
heterocyclic compound that was the only active struc-
ture identified among the substances in the library at
the chosen activity level (30 µM), which otherwise
contained at least 18 other different bis-cationic com-
pounds. In the phosphorylation assay with HPr, this
substance resulted in an EC50 value of 10 µM (Figure
1B) and a value in the same range (21 µM) was obtained
in a kinase assay with Crh (Table 1). Since BSA has
been reported to significantly increase the EC50 value
for inhibitors of tyrosine kinase c-Src,38 an experiment
was performed with an assay containing HPr/Crh and
the bis-cationic compound but without BSA. The doses
required to produce the EC50 were found to be compa-
rable (10 and 11 µM for HPr, 21 and 19 µM for Crh) for
the assays with and without BSA. The enzyme was also
preincubated with the inhibitor for 1 h at 37 °C, and
then the reaction was initiated by adding HPr. The
obtained EC50 value was the same as that received
without preincubation, consistent with a rapid equilib-
rium enzyme system.
ATP Com p etition Assa y a n d Syn er gism /An ta go-
n ism An a lysis. 2a was further investigated for its
ability to compete directly with ATP. Increasing con-
centrations of ATP, up to 6-fold, were included in a

2268 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9
Ramstro¨m et al.
Ta ble 2. Importance of the Heterocyclic Moiety for HPrK/P
Inhibitory Activity in a Kinase Assay with HPr
F igu r e 2. Median-effect plot of the dose-effect relationship
of phosphorylation of HPr. The log(φ) function (which is log-
(fraction phosphorylated/fraction nonphosphorylated)) was
plotted against log conc (∇) (which are the concentrations of
individual compounds (when mixed together the concentration
of the bis-cationic compound is shown in the graph)): 2a alone
(9); ADP alone (O); 2a plus ADP (0). Parallel plots of the
mixtures with respect to the parent compounds indicate that
the inhibition of the phosphorylation of 2a and ADP is
mutually exclusive.
Distin ction of In h ibitor Typ e a n d In ter d om a in
In ter a ction . To distinguish between different types of
inhibition, 2a was included at three different fixed
concentrations (equal to its EC50 value and the half and
the double of this concentration) in an assay with HPr/
Crh at protein concentrations ranging from 0.1 to 200
µM. Nonlinear regression analysis indicated that the
substance is a competitive inhibitor, since the values
for Vmax were constant and those for K0.5 were increased.
To further evaluate the inhibition, the substance was
also tested for potential interface properties because
HPrK/P of B. subtilis exists as a homo-oligomeric
enzyme with a strong positive cooperativity for FBP,
ATP, and other nucleotides.17,18 In an assay with HPr
present at different concentrations of ATP, ranging from
1 µM to 1 mM, the inhibitor was tested at three different
constant concentrations, i.e., the concentration equal to
its EC50 value, a lower concentration, which caused only
a small inhibition, and a concentration that gave
maximal inhibition. The Hill coefficient did not change
(around 4), consistent with no change in cooperativity,
and thus suggests that 2a does not seem to participate
in interface interactions between the different subunits.
P h osp h a ta se Assa y. Since HPrK/P is a bifunctional
enzyme that in addition to its kinase activity also
possesses phosphatase activity, the inhibitor was added
at different concentrations together with a final con-
centration of 1 mM inorganic phosphate to trigger the
phosphatase activity in an assay with HPr/Crh after
complete phosphorylation. The three highest concentra-
tions of 2a together with inorganic phosphate, prepared
in advance, were excluded when the EC50 value was
determined because precipitation was observed after
adding the solutions to the assay mix. Compound 2a
inhibited dephosphorylation of HPr(Ser-P) and Crh(Ser-
P) with an estimated EC50 value of 16 and 14 µM,
respectively (Figure 1C, Table 1).
a
b
A, bromide; B, oxalate. CHN analysis.
phosphorylation assay with 2a at a concentration equal
to its EC50 value. Since the nonspecific binding of the
phosphate donor increased with increasing concentra-
tions of ATP, the values for the samples containing no
enzyme (blank samples) were subtracted from the
samples containing enzyme. The degree of phosphory-
lation did not change with increasing concentrations of
ATP, suggesting that the substance was not competing
with the binding of ATP substrate under these assay
conditions. To further support the nonrecognition of the
ATP site by the bis-cationic compound, ADP was used
as an ATP competitor and synergism/antagonism analy-
sis was performed. The EC50 values for ADP alone in
an assay with HPr was determined to be 4 mM (Table
2). A competition assay with the bis-cationic compound
in combination with ADP was performed at a constant
ratio of the concentrations equal to the ratio of the
compounds at their EC50 values when tested alone. The
concentrations obtained for each compound at EC50
when combined were 5 µM 2a and 2 mM ADP (Table
1). The data were graphed with the median-effect plot
of the dose-effect relationship, where log(dose) was
plotted against log(fraction phosphorylated/fraction non-
phosphorylated) for 2a alone, ADP alone, and 2a with
ADP (Figure 2). Parallel plots of the mixture with
respect to the parent substances indicate that the
inhibitory effect of 2a and ADP is mutually exclusive;
thus, R ) 0. The CI value for the combination when R
) 0 was determined to 1.0, which reveals an additive
effect of the two substances in a mixture. Under the
same conditions, using Crh as a protein substrate
instead of HPr, a similar CI value, 1.1, was obtained
(Table 1). These experiments support an action on two
different sites for ADP and the bis-cationic compound.
The inhibition of both phosphorylation and dephos-
phorylation by the bis-cationic compound was coopera-
tive with a Hill coefficient around 2 (Table 1). In
contrast, ADP did not show such cooperativity under
the experimental conditions used.
Hit Op tim iza tion . A series of derivatives of 2a and
structurally related compounds were prepared and
tested at final concentrations of 10 and 100 µM in a
kinase assay using HPr as the protein substrate. Data

Inhibitors of HPr Kinase/ Phosphatase in B. subtilis
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9 2269
Ta ble 3. Correlation between the Linker Length and the EC50 Values for HPrK/P Inhibition by 2a and Its Homologues
EC50 (µM)
mean (95% confidence interval)
Hill slope
mean (95% confidence interval)
compound
Q(CH2)12
HPr
Crh
HPr
Crh
2a
3a
3b
6a
Q
10 (8.6-12)
20 (16-25)
93 (72-120)
29 (27-32)
21 (17-25)
71 (64-78)
210 (170-250)
32 (28-36)
1.8 (1.3-2.2)
1.1 (0.9-1.3)
1.3 (0.9-1.6)
5.3 (2.8-7.8)
2.4 (1.5-3.3)
2.2 (1.8-2.6)
1.3 (1.1-1.6)
6.4 (3.5-9.3)
Q(CH2)8Q
Q(CH2)5Q
Q(CH2)11CH3
Ta ble 4. Effects of N1 Substitution on HPrK Inhibition of 6a
and Its Derivatives
Table 5 deals with the topological exploration along
the alkyl side chain. In particular, the second identical
cationic head was replaced by other various cations (9
and 10), anions (6m , 8), hydrophobic groups (6k and
6l), or various H bond acceptor-donor systems (6n -
s,w -z). For the optimal fit of a novel putative interac-
tion, various homologues were considered with the
length n varying from 1 to 12.
% inhibition at
% inhibition at
Data taken together allowed for the following com-
ments: (i) Among the lipophilic derivatives with short
or intermediate lengths (6h -l), only the compound with
the largest length 6a (n ) 12) retained some significant
potency. (ii) Introducing anions (carboxylic acids 6m ,
8a b) and H bond acceptor-donor systems (amides 6n -
p ,w -y, alcohols 6q-s) led to inactive compounds. Only
the phthalimide 6z and the aniline 10a showed an EC50
of about 10 µM, as found for the starting bis-cation hit
2a . (iii) Introduction of a second, nonheterocyclic cationic
head within the side chain seemed more promising.
m
R
100 µM (SEM), n ) 3 10 µM (SEM), n ) 3
6a
6b
6c
6d
6e
6f
1
2
3
1
1
1
1
H
H
H
90((1)
98((1)
96((1)
23((4)
21((4)
51((8)
95((1)
18((2)
0((2)
38((4)
0((2)
4-CH3
4-Cl
3,4-Cl
4-OCH3
17((4)
9((4)
34((2)
6g
are listed in Tables 2-5. Replacing the benzimidazole
ring in 2a by other heterocycles, including monocycles
(pyridines 2e and 2f, pyridazine 2g, imidazole 2d ) and
bicycles (isoquinoline 2h , benzothiazole 2c), led to
inactive compounds (Table 2). These results emphasized
the critical role played by the benzimidazole ring for
HPrK inhibition. In addition the N-benzyl moiety
seemed to significantly increase the potency (compare
2a with the corresponding N-methyl derivative 2b).
Table 3 highlights the effect of the length of the linker
between the cationic heads and both their potency and
selectivity toward phosphorylation inhibition of HPr and
Crh. Shortening the length progressively led to de-
creased potencies. In particular, the pentylene deriva-
tive 3b was found to be about 10 times less potent than
the reference compound 2a . It is noteworthy that
suppressing the second cationic head, but with the same
lipophilic side chain, led to a compound (6a ) with 3 times
less potency when compared to the reference compound
in a kinase assay with HPr. However, the mode of action
of this monocation 6a is probably different as shown by
the strong difference in their Hill slope coefficients.
Table 4 shows the effect of the N-benzyl moiety
(compounds 6b and 6c) and the ring substituted deriva-
tives (compounds 6d -g). When compared to the refer-
ence monocation 6a , the superior homologue 6c was
found to be more potent. In addition, the presence of a
p-methoxy group on the aromatic ring in compound 6g
may have some beneficial effect, whereas a similar effect
was not found with the methyl group in 6d . This may
reflect some additional H bond interaction with the
oxygen of the methoxy group in 6g.
First, it is noteworthy that other bis-cations resulting
from primary (9a -c) or secondary (10b) or tertiary
amines (10e) with relatively short lengths (n < 6) were
inactive. However, within the series of increased length
of the spacer (n ) 11 or 12), a significant increase in
potency was observed when introducing an aromatic
ring at the NH2 nitrogen (compare 9c with 10a and
10c). Thus, the second symmetrical aminobenzimidazole
ring could be replaced by other cationic systems such
as benzylamine (10b), morpholine (10d ), and N-methyl-
and N-phenylpiperazines (10h and 10g, respectively).
However, the presence of an aromatic ring in the vicinity
of the cationic head may produce some beneficial effects
(better activities for the set of compounds 10c and 10g
when compared to 9c, 10d , 10h , and 10i). However, the
aniline derivative 10a is probably not protonated at pH
7.4 and may participate in a typical H bond interaction,
since it could be hypothesized for the active noncationic
phthalimide 6z. Taken together, biological data high-
lighted the compounds 6z, 10a , 10c, and 10g as the
most potent inhibitors within this series. Finally, the
good activity of the monocationic compound 6z allowed
us to hypothesize that the second cationic moiety
present in 2a is not always needed for HPrK inhibition.
However, the presence of a phenyl ring in the vicinity
of the chemical moiety R′ seems to be favorable for
activity. On the other hand, R′ (amine or amide) served
as an anchor point for introducing a phenyl ring, which
might establish an additional hydrophobic interaction.

2270 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9
Ramstro¨m et al.
Ta ble 5. Inhibitory Activity of 2a and Related Compounds in a Kinase Assay with HPr
a
Not tested, intermediates.
Discu ssion
may compromise the selectivity and increase the likeli-
hood of nonspecific effects.44 For prokaryotes, intracel-
lular concentrations of glycolytic intermediates, e.g.,
ATP, have been measured in Streptococcus lactis.46
Following the addition of glucose to starved cells, the
intracellular levels of ATP increased to around 2 mM.
Thus, a non-ATP inhibitor binding to its site more
weakly than an ATP-competitive inhibitor binding to
the ATP site could be equally potent intracellularly
because the inhibitor in the first case may compete with
micromolar concentrations of protein substrates com-
pared to millimolar levels of ATP in the second case.
From the screening campaign of the in-house chemical
library, a single inhibitor of HPrK/P in B. subtilis was
identified. In contrast to most kinase inhibitors, which
are competitive with ATP,39-41 this substance inhibited
protein kinase activity independently of the ATP con-
centration, consistent with an alternative mechanism
of inhibition. Non-ATP mimetic kinase inhibitors can
be considered more favorably when compared to ATP-
competitive inhibitors because the ATP-binding site is
highly conserved among a large class of protein kinases.
Consequently, many ATP inhibitors may be described
as general rather than specific protein kinase inhibi-
tors,40 and in order to obtain highly selective ATP site-
directed inhibitors, optimization by use of the structural
information in the vicinity of the ATP site of the specific
kinase needs to be further undertaken.40,42,43 Moreover,
ATP-competitive inhibitors and their application to
intact eukaryotic cells may be limited because of nu-
merous factors such as that high intracellular concen-
trations of inhibitor must overcome physiological mil-
limolar levels of ATP.41,44,45 The consequence of the
requirement of high intracellular levels of inhibitors
Compound 2a seems to have a similar effect in vitro
when considering the degree of inhibition of phospho-
rylation and dephosphorylation regarding HPr and Crh,
respectively, even though a tendency for higher concen-
trations of the inhibitor was needed in a kinase assay
with Crh compared to HPr (Table 1). The tendency for
higher concentrations of inhibitor in a phosphorylation
assay with Crh than with HPr was further supported
when the refinement compounds 3a , 3b, and 6a (Table
3) were analyzed. The background may be due to the
different behavior of the proteins in solution, as was

Inhibitors of HPr Kinase/ Phosphatase in B. subtilis
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9 2271
envisaged by NMR analyses indicating that Crh may
exist as a homodimer, whereas HPr is known to be
monomeric in solution.47,48 However, the concentration
was rather high, 0.5 mM, and yielded a distribution of
70-80% monomer and 20-30% dimer.47,48 When elec-
trospray ionization mass measurements at 20 µM were
performed, no dimer of Crh could be detected.17 Never-
theless, if Crh behaves as a dimer under certain
conditions, the consequence may be the need for higher
concentrations of inhibitor. The fact that both the kinase
and the phosphatase activity was affected suggests a
close association between these two opposing activities.
A close association between the kinase and phosphatase
activity indicating the same catalytic site was reported
for HPrK/P of B. subtilis34,49 as well as for HPrK/P of
L. casei.35,50 Furthermore, the bis-cationic inhibitor
suggests a bivalent contact between the compound and
the enzyme, and the length of the linker proved to be a
critical parameter (the distance between the two
guanidines in the extended state of the molecule with
a 12-carbon spacer was estimated to be around 17 Å as
judged by standard bond lengths). When the alkyl linker
was reduced from 12 to 5 carbon atoms (see 2a and 3b
in Table 3), the EC50 value increased 10 times. The two-
site binding model is further supported by the increas-
ing EC50 value observed for monomer 6a and may
contribute to a higher selectivity of the inhibitor for
HPrK/P than for other kinases.
F igu r e 3. Pharmacophoric model of HPrK/P inhibitors: n )
11, 12; R′ ) basic amine (10c, 10g), fairly amine (10a ), amide
(6z).
Ta ble 6. Influence of the Structure of the Heterocyclic
Amidines (Cationic Head) onto the Selective Inhibition of
Specific Kinasesa
The type of receptor-ligand interaction for the bis-
cationic compound 2a remains unknown, but the struc-
ture of the substance indicates that at least one of the
binding sites may be negatively charged and surrounded
by lipophilic substituents. It can also be envisaged that
a π-cation-type interaction might take place between
cations and aromatic systems in HPrK/P, as observed
in other systems.51 The active compounds emphasized
the importance of the N-benzyl-2-aminobenzimidazole
moiety containing at least an eight-carbon alkyl linker.
Moreover, as shown in Tables 4 and 5, the most active
compounds (percent inhibition of more than 40% at 10
µM) present various scaffolds bearing an aromatic ring.
As supported by the activity of monocations 6a and 6z,
at least one binding site involving a typical lipophilic
benzimidazole ring is the minimum required for binding
to HPr K/P.
a
nt: not tested. a: active. na: non active.
of ADP for its binding to HPrK/P of B. subtilis using
HPr as the protein substrate.18 However, this is not in
contradiction to each other because different substances
may behave differently. For the bis-cationic compound,
the inhibition is cooperative with a Hill coefficient
around 2 in both the phosphorylation and the dephos-
phorylation assays. The cooperativity decreased when
the alkyl linker length decreased (3a ,b) but increased
approximately 3 times when one of the benzyl benzimi-
dazole moieties was removed, the monomer (6a ), com-
pared to the parent compound.
The 1-substituted 2-aminobenzimidazoles represented
here by the prototype shown in Figure 3 belong to the
general class of symmetrical heterocyclic amidines
linked through their endocyclic nitrogen. These bis-
cations are fully protonated at pH 7.4 and are known
to present various pharmacological effects (antimicrobi-
als,52 antifungals53). Moreover, they act at different
targets, including receptors (GPCR muscarinic recep-
tors,54 channel receptors55) and enzymes (protein kinase
C56). It is interesting to note that earlier screening of
the same chemical library against another specific
kinase, calmodulin regulated protein kinase (CaMK),57
highlighted another bis-cationic entity (compound 2g)
as shown in Table 6. Moreover, depending on the nature
of the chemical entity yielding the cationic species
(pyridazine or benzimidazole), some selectivity was
observed. The bis-pyridazine 2g and bis-quinoline 2h
did not inhibit the HPr kinase (see Table 2). It is
noteworthy that hit optimization starting from the
bispyridazine 2g for designing more potent CaMK
inhibitors emphasized the possible replacement of the
second pyridazine nucleus in 2g by a noncationic moiety
bearing an aromatic ring.58
The type of inhibition was examined, suggesting that
2a is a competitive inhibitor because the values for Vmax
were constant and the values for K0.5 were increased
when tested at three different concentrations of the
substance. Furthermore, since HPrK/P of B. subtilis is
an oligomeric enzyme with strong cooperativity for
ATP17,18 the possibility of the bis-cationic compound
acting also as an interface inhibitor was investigated
in an assay with different concentrations of ATP and
at three constant concentrations of inhibitor. However,
the Hill coefficient for ATP did not change, indicating
that the bis-cationic compound does not participate in
interface interactions between the different subunits.
2a was also added to a mixture with ADP. For ADP
alone, the Hill coefficient was determined to be 0.9,
indicating that the nucleotide probably does not induce
any cooperativity under the experimental conditions in
an assay with HPr. A positive cooperativity was re-
ported for the N-methylanthraniloyl (Mant) derivative

2272 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9
Ramstro¨m et al.
with water and brine and dried (Na2SO4). After removal of the
solvent under reduced pressure, the residue was chromato-
graphed on silica gel (ethyl acetate/triethylamine, 95/5) to
provide the free bases of 2a and 2b.
Met h od B: P r oced u r e for Com p ou n d s 2c,e-h a n d
3a ,b. Dibromoalkane (1.5 mmol) was added to a solution of
heterocyclic amidine (3 mmol) in 2-butanone (100 mL), and
the resulting solution was heated at 80 °C for 4 days. After
the mixture was cooled, the precipitate was filtered and
washed twice with 2-butanone (2 × 30 mL). Recrystallization
of the product in isopropyl alcohol yielded the dibromides
2c,e-h and 3a ,b.
Meth od C: P r oced u r e for Com p ou n d 2d . A solution of
imidazole (18.5 mmol) in acetone (50 mL) and KOH (55.5
mmol, 3 equiv) was stirred for 30 min at room temperature.
Dibromododecane (6.25 mmol) was then added, and the
mixture was stirred for 4 h. The solution was then concen-
trated in vacuo and partitioned between EtOAc (50 mL) and
water. The organic layer was washed with brine and extracted
with 2 N HCl (50 mL). The aqueous layer was made basic with
a 20% K2CO3 solution (60 mL) and extracted with AcOEt (60
mL). The organic layer was dried over Na2SO4 and was
evaporated in vacuo to give the free base as an oil, which was
purified by flash chromatography on silica gel (ethyl acetate/
triethylamine, 95/5).
In conclusion, starting from a specific bis-cationic hit
derived from 3-benzyl-2-aminobenzimidazole, a first
SAR analysis clearly identified this heterocyclic system
as a specific feature needed for HPrK inhibition. The
second cationic head could be replaced by a noncationic
chemical moiety bearing an aromatic ring. Further
studies will optimize the nature of both the connecting
group R′ and the aromatic system including substituent
effects at the aromatic ring. When compared with recent
studies dealing with CaMK starting from another bis-
cationic hit compound, they may support the existence
of a similar mode of interaction of the heterocyclic
amidines presented here. Because they do not compete
with the ATP binding site, they may provide interesting
challenges for designing potent protein kinase inhibitors
with significant increased selectivity profiles.
Antimicrobial activity associated with aromatic bis-
cationic compounds is well established. An example is
pentamidine, which is clinically used against Pneu-
mocystis carinii. Compounds related to pentamidine and
other bis-cationic compounds have also been reported
to possess antimicrobial effects.53,59 However, the mech-
anism of action, even though DNA binding has been
hypothesized,59 and the potential of these latter com-
pounds for clinical use seem to be unknown. The first
preliminary results indicated that the parent compound,
as well as some of the compounds related to 2a , i.e., 2b,
3a , 6g, 6y, 10c, 10d , and 10g, inhibited growth of B.
subtilis. However, no inhibition was observed when
samples were tested in cultures of E. coli. Because
significant cytotoxicity was observed for the most in-
teresting compounds (ED50 ≈ 1 µM), the series was
abandoned. Since the HPrK/P is involved in the regula-
tory mechanism in the hierarchical use of carbon
sources for growth of bacteria, a better strategy could
be to evaluate the effect of compounds on growth
retardation. Further studies will need to evaluate
whether inhibitors of HPrK/P are valuable therapeutical
agents as antimicrobials.
P r oced u r e for Oxa la te Sa lts. The free base was dissolved
in a small amount of isopropyl alcohol previously heated to
80 °C, and then 2 equiv of oxalic acid dissolved in isopropyl
alcohol was added. The product recrystallized at room tem-
perature.
N,N′-Bis(2-am in o-1-ben zylben zim idazol-3-yl)-1,12-dode-
ca n e Oxa la te (2a ). Compound 2a was prepared from N-ben-
zyl-2-aminobenzimidazole by method A and was obtained as
a white solid (95% yield). Rf ) 0.26 (AcOEt/TEA, 95:5); mp
1
190-192°C; H NMR (200 MHz, DMSO-d6) δ 1.17-1.24 (m,
16H, 4 -CH2-), 1.60 (m, 4H, 2 -CH2-), 3.82 (t, 4H, J ) 6.7,
2 -CH2N), 5.05 (s, 4H, 2 -CH2Ph), 6.80-6.93 (m, 8H, Ar-H),
7.19-7.29 (m, 10H, Ar-H). Anal. (C44H52N6O8) C, H, N.
N,N′-Bis(2-am in o-1-m eth ylben zim idazol-3-yl)-1,12-dode-
ca n e Oxa la te (2b). Compound 2b was prepared from N-meth-
yl-2-aminobenzimidazole by method A (85% yield). Rf ) 0.23
(AcOEt/TEA, 95:5); mp 181-183 °C; 1H NMR (200 MHz,
CDCl3) δ 1.23-1.31 (m, 16H, 4 -CH2-), 1.63-1.74 (m, 4H, 2
-CH2-), 3.34 (s, 6H, -N1CH3), 3.76 (t, 4H, J ) 6.5, 2 -CH2-
N3), 4.55 (br s, 4H, 2 NH2 exchange with D2O), 6.78-6.85 (m,
4H, H5 and H6), 6.92-6.98 (m, 4H, H4 and H7). Anal.
(C32H44N6O8) C, H, N.
Exp er im en ta l Section
N,N′-Bis(2-a m in oben zoth ia zol-3-yl)-1,12-d od eca n e Ox-
a la te (2c). Compound 2c was prepared from 2-aminoben-
zothiazole by method B (85% yield). Rf ) 0.5 (AcOEt/TEA, 95:
1. Ch em istr y. The structure of all the compounds were
checked by NMR on a Bruker AC-200 or DPX-300 Avance
spectrometer. Shifts are given in ppm (δ) with respect to the
TMS signal, and coupling constants (J ) are given in Hz.
Column chromatography was performed on Merck silica gel
60 (230-400 mesh). Melting points were measured with a
Mettler FP62 apparatus. Mass spectra were recorded on a
Perseptive-Biosystem Mariner electrospray (ES) spectrometer.
Elemental analyses (C, H, N) were carried out at a microana-
lytical center (Lyon).
2-Substituted-1-benzylbenzimidazole 5 was obtained using
the established alkylation of 2-aminobenzimidazole 4 with
KOH and arylalkyl bromide. The alkyl bromide 7 was com-
mercially available or was made by standard procedures from
commercially available starting materials and precursors
described in the literature.
Gen er a l P r oced u r e for th e Syn th esis of th e Bis-
Ca tion ic Com p ou n d s 2a -g a n d 3. Meth od A: P r oced u r e
for Com p ou n d s 2a a n d 2b . Dibromododecane (1.5 mmol)
was added to a solution of N-substituted 2-aminobenzimidazole
(3 mmol) in sulfolane (100 mL), and the resulting solution was
heated at 80 °C for 70 h with stirring. After the mixture was
cooled to 0 °C, the precipitate was filtered, washed with
acetonitrile (30 mL) and then with diethyl ether (50 mL), and
recrystallized from isopropyl alcohol. The resulting solid was
poured into a 20% K2CO3 solution (80 mL) and extracted with
ethyl acetate (3 × 50 mL). The combined extracts were washed
1
5); mp 171-173 °C; H NMR (200 MHz, CDCl3) δ 1.25-1.37
(m, 16H, 4 -CH2-), 1.69-1.79 (m, 4H, 2 -CH2-), 3.90 (t, 4H,
J ) 6.5, 2 -CH2N3), 6.74-6.85 (m, 4H, H7 and NH), 6.92-
7.04 (m, 2H, H6), 7.08-7.26 (m, 4H, H4 and H5). Anal.
(C30H38N4O8S2) C, H, N.
N,N′-Bis(im idazol-1-yl)-1,12-dodecan e Oxalate (2d). 80%
yield; Rf ) 0.24 (AcOEt/TEA, 95:5); mp 158-160 °C; 1H NMR
(200 MHz, CDCl3) δ 1.14-1.23 (m, 16H, 4 -CH2-), 1.75 (m,
4H, 2 -CH2-), 3.91 (t, 4H, J ) 7.1, 2 -CH2N1), 6.89 (br s,
2H, H4), 7.04 (br s, 2H, H5), 7.45 (br s, 2H, H2). Anal.
(C22H34N4O8) C, H, N.
N,N′-Bis(2-a m in op yr id in -1-yl)-1,12-d od eca n e Dih yd r o-
br om id e (2e). Compound 2e was prepared from 2-aminopyr-
idine by method B (95% yield). Rf ) 0.4 (AcOEt/TEA, 95:5);
1
mp 230-232 °C; H NMR (200 MHz, DMSO-d6) δ 1.14-1.23
(m, 16H), 1.65 (m, 4H), 4.16 (t, 4H, J ) 7.0, -CH2N1), 6.90 (t,
4H, J ) 6.4, H4), 7.00 (d, 2H, J ) 8,7, H3), 7.87 (t, 2H, J )
7.1, H5), 8.10 (d, 2H, J ) 5.8, H6), 8.51 (br s, 4H, NH2
exchangeable with D2O). Anal. (C22H36Br2N4) C, H, N.
N ,N ′-Bis(2-a m in o-5-p h e n ylp yr id in -1-yl)-1,12-d od e -
ca n e Dih yd r obr om id e (2f). Compound 2f was prepared from
2-aminobenzothiazole by method B (60% yield). Rf ) 0.37
(AcOEt/TEA, 95:5); mp 182-184 °C; 1H NMR (200 MHz,
DMSO-d6) δ 1.14-1.25 (m, 16H, 4 -CH2-), 1.67 (m, 4H, 2

Inhibitors of HPr Kinase/ Phosphatase in B. subtilis
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9 2273
-CH2-), 4.19 (m, 4H, 2 -CH2N1), 7.32 (m, 6H, Ar-H), 7.53
(m, 4H, Ar-H), 7.70 (B/AB, 2H, J AB ) 7.6, H4), 7.90 (A/AB,
2H, J AB ) 7.6, H3), 8.24 (s, 2H, H6), 8.35 (br s, 4H, NH2
exchangeable with D2O). Anal. (C34H44Br2N4) C, H, N.
N,N′-Bis(2-a m in o-b en zocycloh exa [4,5-c]p yr id a zin -1-
yl)-1,12-d od eca n e Dih yd r obr om id e (2g). Compound 2g
was prepared from 2-aminobenzocyclohexa[4,5-c]pyridazine by
method B (65% yield). Rf ) 0.45 (AcOEt/TEA, 95:5); mp 265-
washed three times with 75 mM H3PO4 to remove unreacted
ATP and, finally, once with ethanol. The papers were dried
and transferred to scintillation vials, and the radioactivity was
determined in a scintillation counter.
Au tom a ted P r otein P h osp h or yla tion Assa y. The radio-
active phosphorylation assay with HPr(His)6, described above,
with the addition of different compounds from the in-house
library, mean final concentration of 0.01 g/L (approximately
30 µM), was carried out in 96-well round-bottom plates
(Greiner, Poitiers, France) one column at a time using a
dispensing robot, Biomek 2000. The final concentration of
DMSO used for the dilutions of the compounds in the in-house
library was 10% v/v. The reactions were initiated by the
addition of HPrK/P(Trx-His6-S-tag), and after incubation at
37 °C for 10 min, the reactions were stopped by adding 100
µL of 75 mM H3PO4 to the samples in the wells in the same
order as the reactions were initiated. Subsequently, 90 µL from
each sample was transferred to 96-well P81 filter plates
(Whatman, Maidstone, England) mounted on a manifold and
vacuum was applied for 1.5 min. The papers were washed
three times with 200 µL of 75 mM H3PO4 and the last time
with 200 µL of ethanol. The wells were incubated with the
washing solutions for 3 min, and the solutions were removed
by applying the vacuum for 20 s except for ethanol where the
vacuum was applied for 5 min. The P81 papers were left to
dry overnight. The next day, 25 µL of scintillation solution was
added to each well and the radioactivity was measured with
a Microbeta 1450 (PerkinElmer, Courtaboeuf, France).
1
267 °C; H NMR (200 MHz, DMSO-d6) δ 1.32-1.50 (m, 16H,
4 -CH2-), 1.99 (m, 4H, 2 -CH2-), 3.18-3.30 (m, 8H, 4 -CH2),
4.49 (m, 4H, 2 -CH2N1), 7.52-7.67 (m, 10H, Ar-H and H4),
9.10 (br s, 4H, NH2 exchangeable with D2O). Anal. (C36H46
-
Br2N6) C, H, N.
N,N′-Bis(1-a m in oisoq u in olin -2-yl)-1,12-d od eca n e Di-
h yd r obr om id e (2h ). Compound 2h was prepared from ami-
noisoquinoline by method B (80% yield). Rf ) 0.5 (AcOEt/TEA,
95:5); mp 283-285 °C; 1H NMR (200 MHz, DMSO-d6) δ 1.14-
1.23 (m, 16H, 4 -CH2-), 1.73 (m, 4H, 2 -CH2-), 4.27 (m, 4H,
2 -CH2N2), 7.32 (d, 2H, J ) 6.9, H4), 7.83-7.96 (m, 8H, Ar-
H), 9.66 (d, 2H, J ) 7.0, H3), 9.21 (br s, 4H, NH2 exchangeable
with D2O). Anal. (C30H40Br2N4) C, H, N.
N,N′-Bis(2-a m in o-1-ben zylben zim id a zol-3-yl)-1,5-p en -
ta n e Dih yd r obr om id e (3a ). Compound 3a was prepared
from 2-amino-1-benzylbenzimidazole with 1,5-dibromopentane
by method B (80% yield). Rf ) 0.2 (AcOEt/TEA, 95:5); mp 299-
301 °C; 1H NMR (200 MHz, DMSO-d6) δ 1.51 (m, 2H, -CH2-
), 1.80 (m, 4H, -CH2-CH2-CH2-), 4.23 (t, 4H, J ) 7.1, 2
-CH2N), 5.51 (s, 4H, 2 -CH2Ph), 7.52 (m, 18H, Ar-H), 9.03
(br s, 4H, 2 NH2 exchange with D2O). Anal. (C33H36Br2N6) C,
H, N.
N,N′-Bis(2-a m in o-1-b en zylb en zim id a zol-3-yl)-1,8-oc-
ta n e Dih yd r obr om id e (3b). Compound 3b was prepared
from 2-amino-1-benzylbenzimidazole with 1,8-dibromooctane
by method B (80% yield). Rf ) 0.2 (AcOEt/TEA, 95:5); mp 302
°C (dec); 1H NMR (300 MHz, DMSO-d6) δ 1.28 (m, 9H, 3
-CH2-), 1.69 (m, 4H, 2 -CH2-), 4.16 (t, 4H, J ) 7.2, 2
-CH2N), 5.51 (s, 4H, 2 -CH2Ph), 7.33 (m, 18H, Ar-H), 8.92
(br s, 4H, 2 NH2 exchange with D2O). Anal. (C36H42Br2N6) C,
H, N.
2. P h a r m a cology. P u r ifica tion of HP r (His)6, Cr h (His)6,
a n d HP r K/P (Tr x-His6-S-ta g) fr om B. su btilis. HPr(His)6,
Crh(His)6, and HPrK/P(Trx-His6-S-tag) from B. subtilis were
purified by Ni-NTA affinity chromatography as described
previously.17 The concentration of HPrK/P(Trx-His6-S-tag) was
determined spectrophotometrically using the Bio-Rad protein
assay (Bio-Rad Laboratories, Mu¨nchen, Germany) with bovine
γ globulin as standard, and the concentrations of Crh(His)6
and HPr(His)6 were determined by UV spectrophotometry
using the extinction coefficient for one and two tyrosine
residues, respectively (1500 and 2900 M-1 cm-1). Protein
solutions were stored at -20 °C.
Ra d ioa ctive P h osp h a ta se Assa y. The phosphorylation
assay described above was carried out to first receive phos-
phorylation of HPr(His)6/Crh(His)6. Then 5 µL of inorganic
phosphate (K2HPO4 and KH2PO4, pH 8.0) was added to give a
final concentration of 1 mM, and the reaction mixture was
further incubated at 37 °C for 2 h. Termination of the reaction
and the rest of the steps was performed as described above.
The final concentration of HPr(His)6/Crh(His)6 was 8 µM.
EC50 Cu r ve. Thirteen different consecutively diluted con-
centrations of tested inhibitor were mixed in advance with
inorganic phosphate. After complete phosphorylation, the
mixtures were added to the assay.
Estim a tion of EC50 Va lu es a n d Kin etic P a r a m eter s.
The program GraphPad Prism was adapted for nonlinear
regression analysis to determine EC50 values, the Hill coef-
ficient, K0.5 (which is the half-saturation constant in the Hill
equation), and Vmax
.
Ack n ow led gm en t. H.R. gratefully acknowledges a
grant from the Swedish Academy of Pharmaceutical
Sciences. This work was supported by a grant from the
ACC Microbiologie, Ministe`re de la Recherche.
In -Hou se Ch em ica l Libr a r y. The compounds were stored
in 96-well plates with the first and last column empty of
substances for control samples, i.e., 80 compounds per plate.
The initial concentration of the compounds, in DMSO, was 10
g/L. The mean molecular weight of the compounds is close to
300 g/mol, which gives a mean molar concentration of ap-
proximately 30 mM. The solutions were further diluted with
DMSO using a dispensing robot, Biomek 2000 (Beckman
Coulter, Roissy CDG, France), to a final concentration of 0.1
g/L (mean molar concentration of approximately 300 µM) in
96-well flat bottom plates (Costar, Corning, NY). All dilutions
were stored under argon at 4 °C.
Ra d ioa ctive P r otein P h osp h or yla tion Assa y. In vitro
phosphorylation of HPr(His)6/Crh(His)6 by HPrK/P(Trx-His6-
S-tag) was performed as described previously.17 Briefly, the
standard assay with a final volume of 20 µL contained 50 mM
Tris, pH 8, 5 mM MgCl2, 0.5 mM ATP, 3.3-5 Bq (leading to
200-300 cpm [γ-32P]ATP/pmol ATP) (Amersham Biosciences,
Uppsala, Sweden), 2 mM FBP, 0.1% BSA, 10 µM HPr(His)6/
Crh(His)6, and to initiate the reaction 100 nM HPrK/P(Trx-
His6-S-tag). After incubation at 37 °C for 10 min, the reaction
was terminated by spotting samples on P81 phosphocellulose
paper (Whatman, Maidstone, England), which were then
Su p p or tin g In for m a tion Ava ila ble: Representative ex-
perimental procedures and spectral data for the preparation
and characterization of compounds 6 and 8-10. This material
is available free of charge via the Internet at http://pubs.ac-
s.org.
Refer en ces
(1) Edelman, A. M.; Blumenthal, D. K.; Krebs, E. G. Protein serine/
threonine kinases. Annu. Rev. Biochem. 1987, 56, 567-613.
(2) Hanks, S. K.; Quinn, A. M.; Hunter, T. The protein kinase
family: conserved features and deduced phylogeny of the
catalytic domains. Science 1988, 241, 42-52.
(3) Hunter, T. Protein kinases and phosphatases: the yin and yang
of protein phosphorylation and signaling. Cell 1995, 80, 225-
236.
(4) Krebs, E.; Fischer, E. H. The phosphorylase b to a converting
enzyme of rabbit skeletal muscle. Biochim. Biophys. Acta 1956,
20, 150-157.
(5) Wang, J . Y.; Koshland, D. E., J r. Evidence for protein kinase
activities in the prokaryote Salmonella typhimurium. J . Biol.
Chem. 1978, 253, 7605-7608.
(6) Cozzone, A. J . Protein phosphorylation in prokaryotes. Annu.
Rev. Microbiol. 1988, 42, 97-125.

2274 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9
Ramstro¨m et al.
(7) Reizer, J .; Novotny, M. J .; Hengstenberg, W.; Saier, M. H., J r.
Properties of ATP-dependent protein kinase from Streptococcus
pyogenes that phosphorylates a seryl residue in HPr, a phos-
phocarrier protein of the phosphotransferase system. J . Bacte-
riol. 1984, 160, 333-340.
(8) Deutscher, J .; Saier, M. H., J r. ATP-dependent protein kinase-
catalyzed phosphorylation of a seryl residue in HPr, a phosphate
carrier protein of the phosphotransferase system in Streptococcus
pyogenes. Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 6790-6794.
(9) Deutscher, J .; Pevec, B.; Beyreuther, K.; Kiltz, H. H.; Hengsten-
berg, W. Streptococcal phosphoenolpyruvate-sugar phospho-
transferase system: amino acid sequence and site of ATP-
dependent phosphorylation of HPr. Biochemistry 1986, 25,
6543-6551.
(10) Galinier, A.; Kravanja, M.; Engelmann, R.; Hengstenberg, W.;
Kilhoffer, M. C.; et al. New protein kinase and protein phos-
phatase families mediate signal transduction in bacterial catab-
olite repression. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 1823-
1828.
(11) Reizer, J .; Hoischen, C.; Titgemeyer, F.; Rivolta, C.; Rabus, R.;
et al. A novel protein kinase that controls carbon catabolite
repression in bacteria. Mol. Microbiol. 1998, 27, 1157-1169.
(12) Kravanja, M.; Engelmann, R.; Dossonnet, V.; Blu¨ggel, M.; Meyer,
H. E.; et al. The hprK gene of Enterococcus faecalis encodes a
novel bifunctional enzyme: the HPr kinase/phosphatase. Mol.
Microbiol. 1999, 312, 59-66.
(13) Dossonnet, V.; Monedero, V.; Zagorec, M.; Galinier, A.; Perez-
Martinez, G.; et al. Phosphorylation of HPr by the bifunctional
HPr kinase/P-Ser-HPr phosphatase from Lactobacillus casei
controls catabolite repression and inducer exclusion but not
inducer expulsion. J . Bacteriol. 2000, 182, 2582-2590.
(14) Huynh, P. L.; J ankovic, I.; Schnell, N. F.; Bruckner, R. Char-
acterization of an HPr kinase mutant of Staphylococcus xylosus.
J . Bacteriol. 2000, 182, 1895-1902.
(15) Brochu, D.; Vadeboncoeur, C. The HPr(Ser) kinase of Strepto-
coccus salivarius: purification, properties, and cloning of the
hprK gene. J . Bacteriol. 1999, 181, 709-717.
(16) Postma, P. W.; Lengeler, J . W.; J acobson, G. R. Phospho-
enolpyruvate:carbohydrate phosphotransferase systems of bac-
teria. Microbiol. Rev. 1993, 57, 543-594.
(17) Ramstro¨m, H.; Sanglier, S.; Leize-Wagner, E.; Philippe, C.; Van
Dorsselaer, A.; Haiech, J . Properties and regulation of the
bifunctional enzyme HPr kinase/phosphatase in Bacillus subtilis.
J . Biol. Chem. 2003, 278, 1174-1185.
(18) J ault, J . M.; Fieulaine, S.; Nessler, S.; Gonzalo, P.; Di Pietro,
A.; et al. The HPr kinase from Bacillus subtilis is a homo-
oligomeric enzyme which exhibits strong positive cooperativity
for nucleotide and fructose 1,6-bisphosphate binding. J . Biol.
Chem. 2000, 275, 1773-1780.
(19) Deutscher, J .; Kuster, E.; Bergstedt, U.; Charrier, V.; Hillen,
W. Protein kinase-dependent HPr/CcpA interaction links gly-
colytic activity to carbon catabolite repression in Gram-positive
bacteria. Mol. Microbiol. 1995, 15, 1049-1053.
(20) Weickert, M. J .; Adhya, S. A family of bacterial regulators
homologous to Gal and Lac repressors. J . Biol. Chem. 1992, 267,
15869-15874.
(21) J ones, B. E.; Dossonnet, V.; Kuster, E.; Hillen, W.; Deutscher,
J .; et al. Binding of the catabolite repressor protein CcpA to its
DNA target is regulated by phosphorylation of its corepressor
HPr. J . Biol. Chem. 1997, 272, 26530-26535.
(22) Fujita, Y.; Miwa, Y.; Galinier, A.; Deutscher, J . Specific recogni-
tion of the Bacillus subtilis gnt cis-acting catabolite-responsive
element by a protein complex formed between CcpA and seryl-
phosphorylated HPr. Mol. Microbiol. 1995, 17, 953-960.
(23) Miwa, Y.; Fujita, Y. Promoter-independent catabolite repression
of the Bacillus subtilis gnt operon. J . Biochem. (Tokyo) 1993,
113, 665-671.
(24) Martin-Verstraete, I.; Deutscher, J .; Galinier, A. Phosphorylation
of HPr and Crh by HprK, early steps in the catabolite repression
signalling pathway for the Bacillus subtilis levanase operon. J .
Bacteriol. 1999, 181, 2966-2969.
(25) Wray, L. V., J r.; Pettengill, F. K.; Fisher, S. H. Catabolite
repression of the Bacillus subtilis hut operon requires a cis-
acting site located downstream of the transcription initiation
site. J . Bacteriol. 1994, 176, 1894-1902.
ackA gene requires sequences upstream of the promoter. J .
Bacteriol. 1998, 180, 5961-5967.
(29) Tobisch, S.; Zuhlke, D.; Bernhardt, J .; Stulke, J .; Hecker, M. Role
of CcpA in regulation of the central pathways of carbon
catabolism in Bacillus subtilis. J . Bacteriol. 1999, 181, 6996-
7004.
(30) Kunst, F.; Ogasawara, N.; Moszer, I.; Albertini, A. M.; Alloni,
G.; et al. The complete genome sequence of the Gram-positive
bacterium Bacillus subtilis. Nature 1997, 390, 249-256.
(31) Galinier, A.; Haiech, J .; Kilhoffer, M. C.; J aquinod, M.; Stulke,
J .; et al. The Bacillus subtilis crh gene encodes a HPr-like protein
involved in carbon catabolite repression. Proc. Natl. Acad. Sci.
U.S.A. 1997, 94, 8439-8444.
(32) Galinier, A.; Deutscher, J .; Martin-Verstraete, I. Phosphorylation
of either crh or HPr mediates binding of CcpA to the Bacillus
subtilis xyn cre and catabolite repression of the xyn operon. J .
Mol. Biol. 1999, 286, 307-314.
(33) Haiech, J . Unpublished observations.
(34) Hanson, K. G.; Steinhauer, K.; Reizer, J .; Hillen, W.; Stu¨lke, J .
HPr kinase/phosphatase of Bacillus subtilis: Expression of the
gene and effects of mutations on enzyme activity, growth and
carbon catabolite repression. Microbiology 2002, 148, 1805-
1811.
(35) Monedero, V.; Poncet, S.; Mijakovic, I.; Fieulaine, S.; Dossonnet,
V.; et al. Mutations lowering the phosphatase activity of HPr
kinase/phosphatase switch off carbon metabolism. EMBO J .
2001, 20, 3928-3937.
(36) Breithaupt, H. The new antibiotics. Nat. Biotechnol. 1999, 17,
1165-1169.
(37) Chou, T.-C.; Talalay, P. Applications of the median-effect
principle for the assessment of low-dose risk of carcinogens and
for the quantitation of synergism and antagonism of chemo-
therapeutic agents. In New Avenues in Developmental Cancer
Chemotherapy; Connors, T. A., Ed.; Academic Press: New York,
1987; pp 37-64.
(38) Maly, D. J .; Choong, I. C.; Ellman, J . A. Combinatorial target-
guided ligand assembly: identification of potent subtype-selec-
tive c-Src inhibitors. Proc. Natl. Acad. Sci. U.S.A. 2000, 97,
2419-2424.
(39) Gray, N. S.; Wodicka, L.; Thunnissen, A. M.; Norman, T. C.;
Kwon, S.; et al. Exploiting chemical libraries, structure, and
genomics in the search for kinase inhibitors. Science 1998, 281,
533-538.
(40) Ono-Saito, N.; Niki, I.; Hidaka, H. H-series protein kinase
inhibitors and potential clinical applications. Pharmacol. Ther.
1999, 82, 123-131.
(41) Meijer, L.; Leclerc, S.; Leost, M. Properties and potential
applications of chemical inhibitors of cyclin-dependent kinases.
Pharmacol. Ther. 1999, 82, 279-284.
(42) Garcia-Echeverria, C.; Traxler, P.; Evans, D. B. ATP site-directed
competitive and irreversible inhibitors of protein kinases. Med.
Res. Rev. 2000, 20, 28-57.
(43) Toledo, L. M.; Lydon, N. B.; Elbaum, D. The structure-based
design of ATP-site directed protein kinase inhibitors. Curr. Med.
Chem. 1999, 6, 775-805.
(44) Penn, R. B.; Parent, J . L.; Pronin, A. N.; Panettieri, R. A., J r.;
Benovic, J . L. Pharmacological inhibition of protein kinases in
intact cells: antagonism of beta adrenergic receptor ligand
binding by H-89 reveals limitations of usefulness. J . Pharmacol.
Exp. Ther. 1999, 288, 428-437.
(45) MacKintosh, C.; MacKintosh, R. W. Inhibitors of protein kinases
and phosphatases. Trends Biochem. Sci. 1994, 19, 444-448.
(46) Mason, P. W.; Carbone, D. P.; Cushman, R. A.; Waggoner, A. S.
The importance of inorganic phosphate in regulation of energy
metabolism of Streptococcus lactis. J . Biol. Chem. 1981, 256,
1861-1866.
(47) Favier, A.; Brutscher, B.; Blackledge, M.; Galinier, A.; Deutscher,
J .; et al. Solution structure and dynamics of Crh, the Bacillus
subtilis catabolite repression HPr. J . Mol. Biol. 2002, 317, 131-
144.
(48) Penin, F.; Favier, A.; Montserret, R.; Brutscher, B.; Deutscher,
J .; et al. Evidence for a dimerisation state of the Bacillus subtilis
catabolite repression HPr-like protein, Crh. J . Mol. Microbiol.
Biotechnol. 2001, 3, 429-432.
(49) Galinier, A.; Lavergne, J . P.; Geourjon, C.; Fieulaine, S.; Nessler,
S.; et al. A new family of phosphotransferases with a P-loop
motif. J . Biol. Chem. 2002, 16, 16.
(26) Dahl, M. K.; Hillen, W. Contributions of XylR, CcpA and HPr to
catabolite repression of the xyl operon in Bacillus subtilis. FEMS
Microbiol. Lett. 1995, 132, 79-83.
(27) Presecan-Siedel, E.; Galinier, A.; Longin, R.; Deutscher, J .;
Danchin, A.; et al. Catabolite regulation of the pta gene as part
of carbon flow pathways in Bacillus subtilis. J . Bacteriol. 1999,
181, 6889-6897.
(28) Turinsky, A. J .; Grundy, F. J .; Kim, J . H.; Chambliss, G. H.;
Henkin, T. M. Transcriptional activation of the Bacillus subtilis
(50) Fieulaine, S.; Morera, S.; Poncet, S.; Monedero, V.; Gueguen-
Chaignon, V.; et al. X-ray structure of HPr kinase: a bacterial
protein kinase with a P-loop nucleotide-binding domain. EMBO
J . 2001, 20, 3917-3927.
(51) Dougherty, D. A. Cation-pi interactions in chemistry and
biology: a new view of benzene, Phe, Tyr, and Trp. Science 1996,
271, 163-168.
(52) Bailey, D. M.; DeGrazia, C. G.; Hoff, S. J .; Schulenberg, P. L.;
O’Connor, J . R.; et al. Bispyridinamines: a new class of topical

Inhibitors of HPr Kinase/ Phosphatase in B. subtilis
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 9 2275
antimicrobial agents as inhibitors of dental plaque. J . Med.
Chem. 1984, 27, 1457-1464.
(53) Del Poeta, M.; Schell, W. A.; Dykstra, C. C.; J ones, S.; Tidwell,
R. R.; et al. Structure-in vitro activity relationships of penta-
midine analogues and dication-substituted bis-benzimidazoles
as new antifungal agents. Antimicrob. Agents Chemother. 1998,
42, 2495-2502.
(54) Ilien, B.; Franchet, C.; Bernard, P.; Weill, C.; Bourguignon, J .
J .; et al. Fluorescence resonance energy transfer to probe human
M1 muscarinic receptor structure and identify biologically active
compounds. Submitted for publication.
(55) Galanakis, D.; Ganellin, C. R.; Malik, S.; Dunn, P. M. Synthesis
and pharmacological testing of dequalinium analogues as block-
ers of the apamin-sensitive Ca(2+)-activated K+ channel: varia-
tion of the length of the alkylene chain. J . Med. Chem. 1996,
39, 3592-3595.
(56) Qin, D.; Sullivan, R.; Berkowitz, W. F.; Bittman, R.; Rotenberg,
S. A. Inhibition of Protein Kinase CR by Dequalinium Ana-
logues: Dependence on Linker Length and Geometry. J . Med.
Chem. 2000, 43, 1413-1417.
(57) Watterson, D. M.; Mirzoeva, S.; Guo, L.; Whyte, A.; Bourguignon,
J . J .; et al. Ligand modulation of glial activation: cell permeable,
small molecule inhibitors of serine-threonine protein kinases
can block induction of interleukin
1 beta and nitric oxide
synthase II. Neurochem. Int. 2001, 39, 459-468.
(58) Mirzoeva, S.; Sawkar, A.; Zasadzki, M.; Guo, L.; Velentza, A.
V.; et al. Discovery of a 3-amino-6-phenyl-pyridazine derivative
as a new synthetic antineuroinflammatory compound. J . Med.
Chem. 2002, 45, 563-566.
(59) Del Poeta, M.; Schell, W. A.; Dykstra, C. C.; J ones, S. K.; Tidwell,
R. R.; et al. In vitro antifungal activities of a series of dication-
substituted carbazoles, furans, and benzimidazoles. Antimicrob.
Agents Chemother. 1998, 42, 2503-2510.
J M021043O