Archives of Oral Biology (2006) 51, 1015—1023

www.intl.elsevierhealth.com/journals/arob

Antimicrobial actions of benzimidazoles against

the oral anaerobes Fusobacterium nucleatum
and Prevotella intermedia
Jiangyun Sheng a, Phuong T.M. Nguyen a,
Jeremiah D. Baldeck a, Jan Olsson b, Robert E. Marquis a,*

a
Department of Microbiology and Immunology and Center for Oral Biology,
The University of Rochester Medical Center, Rochester, NY 14642-8672, USA
b
Institute of Odontology, Göteborg University, S-41390, Göteborg, Sweden

Accepted 8 May 2006

KEYWORDS Summary
Benzimidazoles;
Fusobacterium Background/objective: Benzimidazoles are widely used as proton-pump inhibitors to
nucleatum; control stomach hyperacidity and have been found also to have antimicrobial actions
Prevotella intermedia; against Helicobacter pylori and oral streptococci. Our primary aim was to determine if
Catabolism; they are active also against oral anaerobes associated with gingivitis. Our major focus
Gingivitis was on catabolism because it leads to production of inflammatory metabolites such as
butyrate and ammonia. The benzimidazoles are effective in the protonated form at
acid pH values and cause irreversible inhibition of enzymes associated with formation
of drug-target disulfide bonds.
Methods: Fusobacterium nucleatum ATCC 25586 and Prevotella intermedia ATCC
25611 were grown anaerobically in suspension cultures, harvested, washed and
exposed to the benzimidazole lansoprazole at pH values of 4 or 5 before being
washed and used for standard assays to detect inhibition of catabolic functions,
uptake of the agent and lethality.
Results: Lansoprazole was found to be a bacteriostatic, multi-target antimicrobial
against F. nucleatum under anaerobic conditions inhibitory for amino acid fermenta-
tion and also for glycolysis of glucose or fructose. ID50 values for fermentation of
amino acids and dipeptides by F. nucleatum ranged from 0.05 mM for lysine to
0.25 mM for serine. Fructose catabolism was highly sensitive with an ID50 value of
0.03 mM apparently related to high sensitivity of the phosphoenolpyruvate:fructose
phosphotransferase system, while the ID50 for glucose catabolism by intact cells was

* Corresponding author. Tel.: +1 585 275 1674; fax: +1 585 473 9573.
E-mail address: Robert_Marquis@urmc.rochester.edu (R.E. Marquis).

0003–9969/$ — see front matter # 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.archoralbio.2006.05.002
1016 J. Sheng et al.

some 0.07 mM. Fermentation of aspartate or aspartylaspartate by P. intermedia was

found to be lansoprazole-sensitive with ID50 values of about 0.18 and 0.20 mM,
respectively.
Conclusion: Catabolism of amino acids, dipeptides and sugars by oral anaerobes
associated with gingivitis are sensitive to the inhibitory actions of lansoprazole. Thus,
catabolic pathways are potential targets for use of benzimidazoles against bacteria
involved in gingivitis.
# 2006 Elsevier Ltd. All rights reserved.

Introduction the teeth tends to be alkaline because of the ammo-

Benzimidazoles are widely used as proton-pump
inhibitors to control hyperacidity of the human
stomach, which can lead to ulcers, or in the
extreme, to stomach cancer.1 The inhibitory actions
of benimidazoles require their protonation at acid
pH values leading to formation of a sulfenamide,
which then reacts with protein sulfhydryl groups to
form disulfides with resulting inactivation of sensi-
tive enzymes or other proteins.2 A prime target for
the agents is the proton-pumping (H+ + K+) P-ATPase
of stomach parietal cells. The agents are considered
to have highly specific action against this P-ATPase
and are inhibitory at micromolar concentrations for
the isolated enzyme. Benzimidazoles have been
found3 also to have antimicrobial activities against
Helicobacter pylori, the organism considered to be
the etiologic agent for stomach ulcers. Thus, the
benzimidazoles act to reduce stomach damage by
inhibiting the (H+ + K+) P-ATPase and also by acting
selectively to inhibit H. pylori. In addition, benzi-
midazoles have been found to have anti-oxidant
properties,4,5 even though the sulfenamide form
of the agents acts by oxidising sulfhydryl groups.
We found recently6 that benzimidazoles, primarily
lanzoprazole but also omeprazole, which have pKa
values close to 4, also have potent antimicrobial
actions against oral streptococci. A previous report
by Magalhães et al.7 indicated that lanzoprazole is
inhibitory for a proton-translocating P-ATPase of the
Streptococcus mutans. Our findings indicated that
lanzoprazole also has other actions against oral
streptococci with involvement of multiple targets,
notably catabolic enzymes involved in acid or alkali
production and cariogenicity. Benzimidazoles could
possibly have selective toxicity against cariogenic
bacteria in dental plaque because they are active
only at acid pH values, and so would be potent under
cariogenic conditions, when the pH can be as low as
4, but mutans streptococci remain acidogenic.
The oral diseases gingivitis and periodontitis are
associated with overgrowth of oral anaerobes lead-
ing to tissue-damaging inflammatory responses of
the host.8 In periodontal diseases, the pH value in
the periodontal pockets between the gingivae and
nia releasing activities of multiple microbial and
tissue proteases and the catabolism of amino acids
and peptides by periodontal bacteria.9 Therefore,
the currently used benzimidazoles would not to be
expected to be very effective because they require
acid activation. There are higher pKa benzimida-
zoles, which could possibly be effective, for exam-
ple, rabeprazole has a pKa value of 5 rather than 4,
but benzimidazoles with even higher activation pH
values would be desirable for control of periodontal
organisms. However, pH values in phagolysosomes of
phagocytic cells involved in destruction of period-
ontal bacteria are acidic, in fact, as low as 4, and
benzimidazoles could be activated within phagoly-
sosomes to enhance antimicrobial actions.
Gingivitis can be a preamble to periodontal dis-
eases and involves anaerobic bacteria commonly
found in supragingival plaque, for example, Fuso-
bacterium nucleatum and Prevotella intermedia.8
The general view is that not all gingivitis leads to
periodontal disease but that infectious periodontal
disease usually follows gingivitis. F. nucleatum is
considered an important organism in the mouth also
because it acts as an anchor for many organisms that
are secondary invaders of plaque.10 F. nucleatum is
often considered a transition organism between
supragingival and subgingival plaque. It has transi-
tional metabolism in that it can degrade sugars via
the Embden—Meyerhof—Parnas pathway but mainly
ferments amino acids and peptides.11 Thus, it can be
acid producing when sugars are available or alkali
producing when amino acids and peptides are avail-
able. It does not appear itself to have major pro-
teolytic activities and cannot use proteins such as
casein and albumin as catabolites for growth12,13 but
can associate with bacteria such as Porphyromonas
gingivalis, which do have major endopeptidase
activities. It then can catabolise peptides released
through action of endopeptidases of other oral bac-
teria.
In this paper, we show that F. nucleatum is as
sensitive to the benzimidazole lansoprazole as are
oral streptococci and that cell damage occurs anae-
robically. We also carried out a less extensive inves-
tigation with P. intermedia. Our focus was on
Antimicrobial actions of benzimidazoles
catabolism because it seems that the initial inflam-
matory agents in gingivitis are products of catabo-
lism such as ammonia14 and organic acids, including
butyrate and propionate.15,16
Materials and methods
Bacteria and growth conditions
F. nucleatum ATCC 25586 and P. intermedia ATCC
25611 were stored in 50% glycerol solution at
70 8C, and were grown from storage in a Coy
(Coy Inc., Grass Lake, MI) anaerobic glove chamber
(80% N2, 10% H2 and 10% CO2) on BM agar17 at 37 8C. F.
nucleatum was grown for suspension cultures in
Brain-heart-infusion (BHI) broth (Difco, Detroit, MI)
supplemented with 5 mM glutathione at 37 8C until
the early stationary phase of growth. P. intermedia
was grown similarly except that BM broth was used in
place of BHI. Cells were harvested, washed, and
prepared for experiments under anaerobic conditions
as described previously.18 Viability was assessed by
means of standard most-probable-numbers assays.
Inhibition of fermentation of amino acids/
peptides and of glycolysis
Fermentation of amino acids or dipeptides was
assessed in terms of ammonia production assayed
with commercial ammonia-assay kits (R-Biopharm,
Darmstadt, Germany). Cells in suspensions contain-
ing ca. 1 mg cell dry weight (CDW) per ml in 25 mM
potassium phosphate buffer with 50 mM NaCl and
5 mM MgCl2 were treated with lansoprazole at
desired concentrations for 40 min, or up to 2 h, at
a pH value of 5 to allow for protonation of the agent
and reaction with sensitive targets within cells.
Then the cells were washed to remove unbound
lanzoprazole and suspended in the same buffer at
pH 7. Amino acids (25 mM) or peptides (5 mM) were
added. The substrates tested for F. nucleatum were
glutamate, lysine, serine and glutamylglutamate,
while those for P. intermedia, were aspartate and
aspartylaspartate. At intervals, aliquots of 0.5 ml
were removed, mixed immediately with 0.05 ml of
60% perchloric acid solution, transferred out of the
anaerobic chamber and placed in ice for at least an
hour for product extraction before neutralisation of
acid, dilution and assaying for ammonia.
Cells were prepared in the same manner for assays
of glycolysis. Glucose or fructose was added to give a
final concentration of 10 mM. Disappearance of glu-
cose was assayed with use of glucose oxidase assay
kits purchased from Diagnostic Chemicals Ltd. (Char-
lottetown, P.E.I., Canada). Fermentation of fructose
1017
was assessed with a fructose assay kit from Biotron
Diagnostics Inc. (Hemet, CA).
Uptake of lansoprazole by F. nucleatum
cells
Uptake of lansoprazole was determined by assessing
decreases in absorbance of light of 400 nm wave-
length in supernatant fluids from centrifuged sam-
ples of cell suspensions in salt solution (50 mM KCl
plus 1 mM MgCl2) at specified pH values, as
described previously.6 To determine if lansoprazole
spontaneously degraded under the experimental
conditions, we also included control tubes without
cells. Changes in OD400 of the controls were negli-
gible.
Enzyme assays
Assays of glutamate dehydrogenase and 2-ketoglu-
tarate reductase, the first two enzymes in the 2-
ketoglutarate pathway of glutamate catabolism by
F. nucleatum, were assayed with use of cell extracts
as described by Sheng et al.18 Protein was assayed
for calculating specific activities by the Bradford
method19 with use of the BioRad modification (Bio-
Rad Laboratories Inc., Hercules, CA). Aminopepti-
dase activities in cell extracts of control and lanso-
prazole-treated cells were assayed following the
procedure described by Fujimura and Nakamura20
with use of L-x-glutamyl-p-nitroanilide as substrate.
Assays of phosphoenolpyruvate:sugar phospho-
transferase (PTS) were carried out as described by
Nguyen et al.6 with use of lactic dehydrogenase and
NADH to assay pyruvate released from phosphoe-
nolpyruvate in response to addition of fructose to
cell lysates.
Cytoplasmic levels of ATP were assayed following
the procedures described by Robrish et al.21 with use
of Luciferase/Luciferin assay kits from Promega (San
Luis Obispo, CA) and a TD 20.20 luminometer from
Turner Designs (Sunnyvale, CA). ATP was extracted
from cells by placing 0.2-ml suspension samples
directly into boiling Tris—HCl buffer (20 mM Tris—
HCl, pH 7.75, plus 2 mM EDTA) maintained at 100 8C
for 1.5 min before chilling.
Cytoplasmic levels of NAD and NADH were esti-
mated by the so-called recycling method used by
Snoep et al.22
Inhibition of respiration and NADH oxidase
O2 utilisation by F. nucleatum was assayed by use of
an oxygen electrode as described previously, and
NADH oxidase activity in cell extracts was assayed in
terms of decreased absorbance of 340 nm light
1018 J. Sheng et al.

associated with oxidation of added NADH under
aerobic conditions.18
Chemicals
Lansoprazole was obtained from Sigma Chemical Co.
(St. Louis, MO). It was dissolved in ethanol at suffi-
ciently high concentrations so that there were no
ethanol effects in the experiments described.
Results
Inhibition of catabolism
Glutamate is considered a major catabolite for F.
nucleatum in plaque because it makes up more than
50% of the available amino acids.23 Glutamate is
degraded by the 2-ketoglutamate pathway leading
to production of ammonia, acetate and butyrate.24
The data presented in Fig. 1A shows that lanzopra-
zole treatment resulted in reduced capacities of
intact cells to catabolise glutamate, assayed in
terms of ammonia production, with an average
ID50 value of ca. 0.2 mM. For these assays, the cells
were incubated with lansoprazole at pH 5 before
being washed and assayed for ammonia production
from glutamate at pH 7. Incubation of cells at pH 5
for 40 min in the absence of lansoprazole did not
result in diminished capacity to catabolise gluta-
mate. Thus, the assay detected inhibition of gluta-
mate catabolism that was not reversible with simple
washing. Lansoprazole inactivation of glutamate
fermentation could be prevented through use of
sulfhydryl agents such 2-mercapto-ethanol but
could not be reversed by the agents (data not
shown). When untreated cells or those treated with
0.5 or 1.0 mM lansoprazole for 40 min at pH 5 were
washed and used to inoculate new, supplemented

Figure 1 Inhibition by lansoprazole (LAN) of glutamate (A) or glutamylglutamate (B) catabolism of intact cells of
Fusobacterium nucleatum and of aminopeptidase (C), GLDH: glutamate dehydrogenase or OGR: 2-ketoglutarate
oxidoreductase (D) activities in cell extracts. Error bars indicate standard deviations with N = 3 in three separate
experiments. Control activities were 3.68 mmol NH3 mg 1 cell dry weight/2 h for B, 53.42 nanomol p-nitroanilide
released/mg protein/h for (C) and 4.9 (GLDH) or 26.6 (OGR) micromol NAD or NADH metabolised/min/mg protein for (D).
Antimicrobial actions of benzimidazoles
BHI broth with 1% inocula, there was a lag phase of
some 5 h before control exponential growth began,
compared with 22 and 37 h, respectively, for cul-
tures inoculated with cells treated with 0.5 or
1.0 mM lansoprazole. It appears then that lansopra-
zole did not cause DNA damage and gene inactiva-
tion but mainly irreversible enzyme inactivation.
When the cells were allowed to synthesise new
enzyme during growth in complete medium, their
fermentative capacities were restored. The action
of lansoprazole at enzyme-inactivating levels can be
considered bacteriostatic, but as found previously,6
lansoprazole at levels of 0.1 to 0.5 mM can be
bactericidal, as determined by most probable num-
bers analyses, for oral streptococci and F. nucleatum
over time. Still, it seems that the major action of
lansoprazole is bacteriostatic.
Peptides are considered to be important catabo-
lites for oral anaerobes. They are taken up via
peptide transport systems and then hydrolysed in
the cytoplasm by peptidases, notably aminopepti-
dases in F. nucleatum.13 Catabolism of the peptide
glutamyl-glutamate was approximately as sensitive
to lansoprazole inactivation as was catabolism of
glutamate (Fig. 1B), and the average ID50 was ca.
0.15 mM. Again, cells were exposed to lansoprazole
at pH 5 prior to washing and use for pH-7 assays of
ammonia production from glutamyl-glutamate.
Aminopeptidases in the cytoplasm of intact cells
were highly sensitive to inactivation by lansoprazole
(Fig. 1C) with an ID50 value of 0.03 mM. For the assay,
intact cells were first treated with the indicated
levels of lansoprazole at pH 5, centrifuged, washed
and then used to prepare cell extracts for aminopep-
tidase assays. The low value for ID50 for aminopepti-
dase activity was surprising in relation to the ID50
value of 0.15 mM for catabolism of glutamyl-gluta-
mate by intact cells. Possibly, other peptidases can
cleave the dipeptide, although amino peptidases are
considered to be the major peptidases required for
growth of the F. nucleatum, which does not have
readily detectable endopeptidase activities.13 Possi-
bly, extraction of the enzyme from cells augments
damage caused previously by lansoprazole.
Lansoprazole inhibition of glutamate degradation
could be related to inhibition of the second enzyme in
the catabolic pathway, 2-ketoglutarate oxidoreduc-
tase. The data presented in Fig. 1D indicate an ID50
value of about 0.3 mM for the enzyme in cell extracts,
only somewhat higher the ID50 value of 0.2 for inhibi-
tion of glutamate catabolism by intact cells. In con-
trast, the first enzyme in the pathway, glutamate
dehydrogenase was insensitive to lansoprazole under
the experimental conditions used (Fig. 1D). Other
enzymes in the pathway may also be sensitive to
lansoprazole inhibition, but inactivation of 2-keto-
1019
Figure 2 Reduction in cytoplasmic ATP levels of cata-
bolising suspension cells of F. nucleatum following prior
treatment with lansoprazole. Cells were treated with
lansoprazole at pH 5.0 for 40 min before being washed
and given 25 mM glutamate at pH 7.0 without additional
lansoprazole.
glutarate oxidoreductase would be sufficient for inhi-
bition of glutamate fermentation by intact cells.
The overall inhibitory effects of lansoprazole on
catabolism of F. nucleatum were reflected by
decreased capacities of lansoprazole-treated cells
to pool ATP in the cytoplasm after being washed to
remove the agent and then, in this set of experi-
ments, being given glutamate at pH 7.0 (Fig. 2). As
shown, subsequent rises in levels of ATP pools were
reduced by somewhat more than half in cells treated
with 0.1 mM lansoprazole, and this finding correlates
well with the inhibitory effects found for glutamate
catabolism by the cells. Similarly reduced capacities
to produce NADH were found for cells of F. nucleatum
exposed to lansoprazole and then washed before
being given 25 mM glutamate. Net production of
NADH by cells catabolising glutamate over 2 h was
reduced to approximately half that of control cells by
exposure to 0.1 mM lansoprazole at pH 5 for 40 min–—
an average of 0.39 versus 0.86 nmol NADH per mg cell
dry weight. Exposure to 0.5 mM lansoprazole resulted
in essentially complete elimination of any capacity to
produce NADH. Thus, lansoprazole had major effects
on cell energetics in terms of reduced capacities of
the organism to maintain cytoplasmic pools of ATP or
NADH.
Determinations were made of the kinetics of inac-
tivation of glutamate catabolic capacity of intact
cells of F. nucleatum during exposure to 0.25 mM
lanzoprazole at pH 5 for different periods of time
followed by centrifugation, washing and assaying for
ammonia production from glutamate. The control
value for cells not exposed to lansoprazole was
3.41  0.09 mmol NH3 produced per mg cell dry
weight per hour. The values for cells exposed to
1020 J. Sheng et al.

lansoprazole at pH 5 for 10, 20, 30 or 40 min were:

2.26  0.05, 1.64  0.18, 0.64  0.13, 0.38  0.04,
respectively. Thus, fermentative capacity was
reduced by about one third after 10 min exposure,
and by 30 min, inactivation was essentially complete
in terms of ammonia produced specifically in
response to glutamate. Similar kinetics were
obtained for inactivation of glycolytic capacities of
intact cells with 50% inactivation after approximately
ten minutes exposure at pH 5 (data not presented).
F. nucleatum can also catabolise other amino
acids in addition to glutamate, for example, lysine
and serine. As shown by the data of Fig. 3, lanso-
prazole inactivated catabolic systems for lysine and
serine, assayed in terms of ammonia production in
response to addition of substrate, with ID50 values of
about 0.1 mM after incubation with lansoprazole at
pH 5 for 40 min, washing and assaying for ammonia
production from the amino acids over a 2-h incuba-
tion period.
F. nucleatum has limited capacities to catabolise
sugars such as glucose and fructose at only about one
tenth the rate we determined6 for S. mutans, which
is noted for its acidogenicity. F. nucleatum has a PTS
for uptake of fructose but takes up glucose via an
ABC permease system.25 As shown in Fig. 4A, lanso-
prazole was effective for inactivation of glucose
catabolism with an ID50 of about 0.07 mM. Fructose Figure 4 Inhibition by lansoprazole of glucose utilisa-
catabolism by intact cells was also highly sensitive tion (A) and of fructose-PTS activity (B) of intact cells of F.
nucleatum in suspensions. The cells for glycolysis assay
were initially treated with lansoprazole at pH 5 for 40 min,
while those for assay of fructose-PTS activity were treated
with the agent at pH 5 for 1 h before washing and use.
Cells for assay of glycolysis were suspended in Coles29
buffer at pH 7 and given 10 mM glucose. One-hundred
percentage fructose-PTS activity was 0.026 micromol pyr-
uvate formed/mg cell dry weight/h. N = 3, and error bars
indicate standard deviations.

with an ID50 value of about 0.03 mM. Lansoprazole

Figure 3 Lansoprazole inhibition of lysine and serine
catabolism of F. nucleatum over 2 h after prior treatment
with lansoprazole, washing, and addition of 25 mM amino
acids. N = 3, and error bars indicate standard deviations.
inhibition of sugar catabolism was irreversible, even
when cells were treated with agents such as 2-
mercapto-ethanol or reduced glutathione after
initial exposure to lansoprazole. However, inhibition
could be prevented by addition of the reduced
sulfhydryl agents to cells either before or at the
same time as the benzimidazole. The extreme sen-
sitivity of fructose catabolism to lansoprazole could
be related to a high level of sensitivity of the
fructose-PTS (Fig. 4B) assayed with intact cells
exposed to the agent then washed and permeabi-
lised for the PTS assay. The ID50 value is about
0.005 mM, below the ID50 for catabolism of fructose
by intact cells. The fructose-PTS of S. mutans was
found to have an ID50 value of 0.03 mM for lanso-
prazole inhibition (data not shown), which was
Antimicrobial actions of benzimidazoles
somewhat lower than the value of 0.09 mM for
inhibition of the glucose-PTS of the streptococcus.6
Overall, the findings indicate that phosphotransfer-
ase systems are highly sensitive targets for benzi-
midazole inhibition of sugar catabolism.
P. intermedia is another oral anaerobe commonly
associated with gingivitis, and as shown in Fig. 5A,
its catabolic capacity for degradation of aspartic
acid was inhibited by lansoprazole at pH 5 for 40 min
with an ID50 of ca. 0.18 mM, a value close to that
obtained for loss of glutamate catabolic capacity of
F. nucleatum. Aspartate is considered26 to be a
major catabolite for P. intermedia. However, the
organism probably catabolises mainly peptides in
the gingival crevice,27 and as shown by the data of
Fig. 5B, lansoprazole (40 min at pH 5) reduced its
capacity for catabolism of the dipeptide aspartyl-
aspartate with an ID50 value of about 0.20 mM.
Figure 5 Lansoprazole inhibition of catabolism of aspar-
tate (A) or aspartylaspartate (B) by suspension cells of P.
intermedia after prior treatment with lansoprazole for
40 min at pH 5 followed by washing and incubation at pH
7.0 for 2 h to allow for catabolism. The control rate for A
was 1.37 micromol NH3 produced/mg cell dry weight/2 h.
N = 4 for (A) and 3 for (B). Error bars indicate standard
deviations.
1021
Uptake of lansoprazole
Uptake of lansoprazole by F. nucleatum was very
pH dependent, as it was for oral streptococci.6 The
average rate of uptake by F. nucleatum at pH 4.5
from a solution with 0.25 mM lansoprazole was
2.74 mmol/mg cell dry weight/hour, and the value
decreased to 1.30 a pH 5.0 and 0.52 at pH 5.5.
Maximal uptake at each of the chosen pH values
was approximately 3.6 mmol lansoprazole/mg cell
dry weight. Thus, total uptake was extensive and
appeared to involve a large fraction of the
expected total sulfhydryl groups in the cell, assum-
ing that uptake derives entirely from disulfide
bond formation.
Inhibition of respiration
Although F. nucleatum grows only anaerobically, or
at least microaerophilically, it has an active
respiration, and for example, intact cells metabo-
lised an average of 2.7 nmol O2/min/mg cell dry
weight with fructose as substrate for generating
NADH. Much of the respiration of the organism is
thought to involve NADH oxidase.25 Lanzoprazole
inhibited respiration of the organism but with a
high ID50 value of about 0.5 mM and with only some
50% inhibition of total activity. The ID50 value for
NADH oxidase activity of cells extracts of F. nucle-
atum (Fig. 6) was approximately 0.25 mM (pH 5 for
40 min), which is close to that for inhibition of the
benzimidazole-sensitive respiration of intact cells.
Figure 6 Lansoprazole inhibition of NADH oxidase of F.
nucleatum. Cells were exposed to the agent at pH 5 for
40 min before washing and preparation of cell extracts
with use of a Mini-Beadbeater (Bio Spec Products Inc.,
Bartlesville, OK). The cell extracts were then used for
assays of NADH oxidase activity assessed in terms of NADH
oxidation and changes in absorption of light of 340-nm
wavelength.
1022
In cell extracts, NADH oxidase activity could be
completely inhibited by lansoprazole, and so it
appears that the approximately 50% of respiration
inhibited by lansoprazole is probably that due to
NADH oxidase.
Discussion
A major conclusion from our results is that benzi-
midazoles, specifically lansoprazole, have multi-
target, inhibitory actions against the oral anaerobe,
F. nucleatum, which is commonly associated with
gingivitis and serves as an anchor for secondary
colonisers of dental plaque. Lansoprazole was effec-
tive also for inhibiting amino acid and peptide fer-
mentation by P. intermedia, another anaerobe
associated with gingivitis. Oral anaerobes can be
readily detected in supragingival plaque,28 where
they would be sensitive to benzimidazoles during
plaque acidification phases but may also be a source
of infecting organisms for periodontal diseases. Our
major focus in this manuscript was on catabolism
because it seems that catabolic products of the
organisms are important in initiating the inflamma-
tion of gingivitis. Catabolism of sugars, especially
fructose, also was sensitive to inactivation by lan-
soprazole at acid pH values. Aminopeptidase activ-
ity in cell extracts from F. nucleatum was highly
sensitive to lansoprazole inhibition. However, the
high level of sensitivity may be related to the
enzyme being removed from cells and from protec-
tive actions of the cell membrane and cell sulfhydryl
groups. The high sensitivity of fructose catabolism
to lansoprazole was related to high sensitivity of the
fructose-PTS system, even in intact cells. Lansopra-
zole was inhibitory also for aspartate or aspartyl-
aspartate fermentation by P. intermedia, but with
ID50 values between 0.1 and 0.2 mM. It seems that
for benzimidazoles to be useful antimicrobial agents
for oral use, they would have to be applied repeat-
edly and retained. Because uptake of benzimida-
zoles by bacteria involves formation of covalent
disulfide bonds, their actions for enzymes are irre-
versible. However, if the organisms retain any capa-
city to synthesise new proteins, then with time, the
inhibition is reversed as inactivated enzymes are
replaced by new, active enzymes. The situation is
similar in some ways to that in the human stomach
where benzimidazole inhibition of P-ATPases is
slowly reversed, and there is need for daily intake
of the agent to control hyperacidity. Lansoprazole
can be cidal in killing assays. Moreover, the agent
can work in conjunction with other agents to bring
about killing, However, it seems that killing requires
more extensive damage than that for the enzyme
J. Sheng et al.
inactivation described in this manuscript, and we
consider lansoprazole inhibition to be mainly bac-
teriostatic.
F. nucleatum and P. intermedia did not appear to
be much more sensitive to lansoprazole than are oral
streptococci, although the anaerobes and strepto-
cocci have both different and shared targets for
damage. Thus, it appears that benzimidazoles would
not be highly selective against oral anaerobes com-
pared with facultative organisms. In fact, selective
toxicity would be more likely against highly acido-
genic organisms because the agent is active only after
protonation. However, our data do show that lanso-
prazole is an effective inhibitor under anaerobic
conditions. The total uptake of lansoprazole by F.
nucleatum at pH 4.5, some 3.6 mmol/mg cell dry
weight under substrate-saturating conditions, and
the value suggests that a large fraction of the sulfhy-
dryl groups in the organism can react with the agent,
assuming that cells are approximately 50% protein on
a dry weight basis. The kinetics of uptake were
affected markedly by pH, but total uptake was found
to be largely independent of the rate of uptake.
Presumably, the greater time required for uptake
at higher pH values is related to the requirement
for lansoprazole to be protonated before it can react
to form disulfides with cell targets. More superficial
enzymes such as the membrane-located components
of the PTS would likely react more rapidly with the
agent, whereas inactivation of cytoplasmic targets
such as 2-ketoglutamate oxidoreductase would
require penetration of the membrane by the agent
before reaction with sulfhydryl groups of the enzyme.
Again, it seems that for benzimidazoles to be effec-
tive against gingivitis, exposure to the agents would
have to be frequent in association with toothpaste or
a mouthwash.
The results of our studies with benzimidazoles
indicate that they affect multiple targets in bac-
teria, including oral anaerobes, and that their anti-
microbial actions are not confined to inhibition of P-
type ATPases. An important aspect of the actions of
the agents is that they require acid conditions for
activity. Therefore, they would not be harmful at
higher pH values but would be activated only when
acidified, for example, in carious dental plaque, in
early stages of gingivitis or in the acidified lumen of
phagolysosomes.
Acknowledgments
This work was supported by awards R01 DE 13683
and R01 DE 06127 from the National Institute of
Dental and Craniofacial Research of the U.S. Public
Health Service.
Antimicrobial actions of benzimidazoles
References
1. Olbe L, Carlsson E, Linberg P. A proton-pump inhibitor expe-
dition: the case history of omeprazole and esomeprazole.
Nature Rev 2003;2:132—9.
2. Im WB, Sih JC, Blakeman DP, McGrath JP. Omeprazole, a
specific inhibitor of gastric (H+ K+)-ATPase, is a H+-acti-
vated oxidising agent of sulfhydryl groups. J Biol Chem
1985;260:4591—7.
3. Iwahi T, Satoh H, Nakao M, Iwasaki T, Yamazaki T, Kubo K,
et al. Lansoprazole, a novel benimidazole proton pump
inhibitor, and its related compounds have selective activity
against Helicobacter pylori. Antimicrob Agents Chemother
1991;35:490—6.
4. Lapenna D, de Gioia S, Ciofani G, Festi D, Cuccurullo F.
Antioxidant properties of omeprazole. FEBS Lett 1996;382:
189—92.
5. Biswas K, Bandyopadhyay U, Chattopadhyay I, Varadaraj A,
Ali E, Banerjee RK. A novel antioxidant and antiapoptotic role
of omeprazole to block gastric ulcer through scavenging of
hydroxyl radical. J Biol Chem 2003;278:10993—1001.
6. Nguyen PTM, Baldeck JD, Olsson J, Marquis RE. Antimicrobial
actions of benzimidazoles against oral streptococci. Oral
Microbiol Immunol 2005;20:93—100.
7. Magalhães PP, Pauline TP, Thedei Jr J, Larson RE, Ciancaglini
P. A 100 kDa vanadate and lanzoprazole-sensitive ATPase
from Streptococcus mutans membrane. Arch Oral Biol
2003;48:815—24.
8. Moore WEC, Holdeman LV, Cato EP, Smibert RM, Burmeister
JA, Best AM, et al. Bacteriology of human gingivitis. J Dent
Res 1987;66:989—95.
9. Marsh P, Martin MV. Oral microbiology. Oxford: Butterworth-
Heinemann; 1999.
10. Kolenbrander PE, Andersen RN, Kazmerzak KM, Wu R, Palmer
Jr RJ. Spatial organisation of oral bacteria in biofilms. Meth
Enzymol 1999;310:322—32.
11. Rogers AH, Gully NJ, Pfenning AL, Zilm PS. The breakdown
and utilisation of peptides by strains of Fusobacterium nucle-
atum. Anaerobe 1998;4:111—6.
12. Rogers AH, Zilm PS, Gully NJ, Pfenning AL, Marsh PD. Aspects
of the growth and metabolism of Fusobacterium nucleatum
ATCC 10953 in continuous culture. Oral Microbiol Immunol
1991;6:250—5.
13. Rogers AH. Studies of fusobacteria associated with period-
ontal diseases. Aust Dent J 1998;43:105—9.
14. Burne RA, Chen Y-YM. Bacterial ureases in infectious dis-
eases. Microbiol Infect 2000;2:533—42.
15. Bartold PM, Gully NJ, Zilm PS, Rogers AH. Identification of
components in Fusobacterium nucleatum chemostat-culture
1023
supernatants that are potent inhibitors of human gingival
fibroblast proliferation. J Periodontal Res 1991;26:314—22.
16. Jeng J-H, Chan C-P, Ho Y-S, Lan W-H, Hsieh C-C, Chang M-C.
Effects of butyrate and propionate on the adhesion, growth,
cell cycle kinetics and protein synthesis of cultured human
gingival fibroblasts. J Periodontol 1999;70:1435—42.
17. Shah HN, Williams RAD, Bowden GH, Hardie JM. Comparison
of the biochemical properties of Bacteroides melaninogen-
icus from human dental plaque and other sites. J Appl
Bacteriol 1976;41:473—92.
18. Sheng J, Nguyen PTM, Marquis RE. Multi-target antimicrobial
actions of zinc against oral anaerobes. Arch Oral Biol
2005;50:747—57.
19. Bradford MM. A rapid and sensitive method for the quantita-
tion of microgram quantities of protein utilising the principle
of protein-dye binding. Anal Chem 1976;72:248—54.
20. Fujimura S, Makamura T. Isolation and characterisation of a
protease from Bacteroides gingivalis. Infect Immunol
1987;55:716—20.
21. Robrish SA, Kemp CW, Bowen WH. Use of extractable ade-
nosine triphosphate to estimate the viable cell mass in dental
plaque samples obtained from monkeys. Appl Environ Micro-
biol 1978;35:743—9.
22. Snoep JL, Teixeira de Mattos MJ, Postma PW, Neijssel OM.
Involvement of pyruvate dehydrogenase in product forma-
tion in pyruvate-limited anaerobic chemostat cultures of
Enterococcus faecalis NCTC 775. Arch Microbiol 1990;154:
50—5.
23. Singer DL, Kleinberg I. The free amino acids in human dental
plaque. Arch Oral Biol 1983;28:873—8.
24. Gharbia SE, Shah HN. Pathways of glutamate catabolism
among Fusobacterium species. J Gen Microbiol 1991;137:
1201—6.
25. Kapatral V, Anderson I, Ivanova N, Reznik G, Los T, Lykidis A,
et al. Genome sequence and analysis of the oral bacterium
Fusobacterium nucleatum strain ATCC 25586. J Bacteriol
2002;184:2005—18.
26. Takahashi N, Yamada T. Pathways for amino acid metabolism
by Prevotella intermedia and Prevotella nigrescens. Oral
Microbiol Immunol 2000;15:96—102.
27. Takahashi N, Sato T. Dipeptide utilisation by the periodontal
pathogens Porphyromonas gingivalis. Prevotella intermedia,
Prevotella nigrescens and Fusobacterium nucleatum. Oral
Microbiol Immunol 2002;15:50—4.
28. Gmur R, Marinello CP, Guggenheim B. Periodontitis asso-
ciated bacteria in supragingival plaque of dental hygienists:
stability of carrier state and clinical development. Eur J Oral
Sci 1999;107:225—8.
29. Coles Jr RS. Glucose utilisation by resting cells of Fusobac-
terium polymorphum. Arch Oral Biol 1977;22:87—90.