JAC
Journal of Antimicrobial Chemotherapy (1997) 40, 659–666
Effects of triclosan and triclosan monophosphate on maximum specific
growth rates, biomass and hydrolytic enzyme production of
Streptococcus sanguis and Capnocytophaga gingivalis in continuous
culture
J. Greenmana*, C. McKenziea and D. G. A. Nelsonb
a
Bristol Oral Microbiology Unit, Faculty of Applied Sciences, University of the West of England,
Frenchay Campus, Coldharbour Lane, Bristol BS16 1QY, UK; bOral Care Technology Division,
Miami Valley Laboratories, Procter & Gamble Co., PO Box 398707, Cincinnati, OH 45239, and
Present address: Consumer Health Care Group, Pfizer Inc., 400 Welbro Road, Parsippany, NJ 07054, USA
Dental plaque species, Streptococcus sanguis and Capnocytophaga gingivalis, were grown in
continuous culture with progressively increasing concentrations of triclosan or its phosphorylated derivative, triclosan monophosphate (TMP). For both organisms, the maximum
specific growth rates decreased with increasing concentrations of triclosan or TMP until
complete inhibition of growth occurred, which for S. sanguis was at 20 mg/L and 50 mg/L, and
for C. gingivalis was at 10 mg/L and 5 mg/L for triclosan and TMP respectively. For both
species, biomass levels remained approximately constant or, in some cases, increased
slightly at low levels of triclosan or TMP. However, biomass levels then decreased significantly as the triclosan or TMP concentrations approached lethal levels. For S. sanguis, levels
of hydrolytic enzymes (acid phosphatase, leucine aminopeptidase and esterase) generally
remained approximately constant or increased with increasing concentrations of triclosan
or TMP until close to inhibitory levels where enzyme levels were reduced. The ratio of extracellular soluble enzmyes to cell-bound enzymes remained constant or increased slightly with
increasing levels of triclosan or TMP. For C. gingivalis, production of hydrolytic enzymes
(neutral phosphatase, leucine aminopeptidase and trypsin-like protease) remained constant or
were reduced when grown with low levels of triclosan and TMP but in some cases increased
with higher levels of agents. The proportion of extracellular soluble activity increased
significantly when concentrations of agent neared inhibitory levels. The results taken together
show that the physiology of cells is significantly altered and that hydrolytic enzymes are
released from the cells when these are grown in the presence of increasing concentrations of
triclosan or TMP. Enzyme release is more pronounced in the Gram-negative C. gingivalis and
indicates that triclosan or TMP can cause membrane perturbation with subsequent release of
membrane-located (S. sanguis) or periplasmic (C. gingivalis) hydrolytic enzymes. S. sanguis
was more sensitive to triclosan than TMP while C. gingivalis was more sensitive to TMP. This
suggests that, in C. gingivalis, TMP may diffuse into the cell wall more easily than triclosan
and then be converted to triclosan by phosphatase activity within the cell wall complex, where
it may give rise to high localized concentrations and subsequent cell damage.
Introduction
The antimicrobial agent triclosan (2,4,4 -trichloro-2 hydroxydiphenyl ether) is used in a number of oral hygiene
products. It can inhibit the growth of a wide range of
microorganisms, including Gram-positive and Gramnegative bacteria and fungi, with MICs usually ranging
from 0.1 to 30 mg/L.1,2 The mechanisms of action of
triclosan are not entirely clear. Due to its hydrophobic and
lipophilic nature, triclosan adsorbs to the lipid portion of
*Tel: +44-(0)117-9763836; Fax: +44-(0)117-9763871.
659
© 1997 The British Society for Antimicrobial Chemotherapy
J. Greenman et al.
the bacterial cell membrane3 where it is thought to act. One
of the main disadvantages of triclosan is its relatively low
solubility in aqueous solutions. Triclosan monophosphate
(TMP) is a phosphorylated derivative of triclosan which, in
comparison, is highly soluble in aqueous solutions. It is
thought that, within the oral environment, TMP (which
itself may be devoid of antimicrobial activity) is hydrolysed
to triclosan by microbial phosphatases.4
The use of continuous culture to study the effects of antimicrobial compounds on the growth and physiology of
particular microorganisms is now well established.5–7 In
batch culture, all parameters (e.g. pH, Eh (redox potential),
substrate concentration, growth rate, cell numbers) change
with time and the ratio of antimicrobial molecules to target
sites (cells) changes rapidly as cells multiply. Either the
antimicrobial compound is present at a high enough concentration to inhibit growth completely, or growth occurs
and any subsequent sub-lethal effects of antimicrobial on
the cell are rapidly masked or ‘diluted out’ due to the exponential increase in cell number. The advantage of using
continuous culture over batch culture is that during steadystate growth in continuous culture the environment
remains constant, so the multiplication rate of cells and the
ratio between biomass (and therefore target sites) and antimicrobial concentration do not fluctuate. Thus, it can be
seen that continuous culture using a chemostat is the only
practical means by which sub-lethal effects of antimicrobial
compounds on cells may be studied.
In the present study, the production and release of
hydrolytic enzymes (acid phosphatase, leucine aminopeptidase and esterase for Streptococcus sanguis; neutral phos phatase, leucine aminopeptidase and trypsin-like protease
for Capnocytophaga gingivalis) were used as measures or
indicators of microbial stress or membrane perturbation in
cells exposed to sub-lethal challenge by triclosan and TMP.
For antimicrobial agents that perturb cell membrane functions it would be expected that such membrane-bound or
secreted hydrolytic enzymes would show changes in their
patterns of production and release when exposed to
increasing levels of antimicrobial agent. Measurement of
phosphatase activity is of particular interest since this
enzyme is likely to be of significance in the conversion of
TMP to active triclosan by hydrolysis.4
The current series of experiments was designed to
compare TMP with triclosan at a range of sub-lethal
concentrations in the growth medium in continuous culture
for two species of dental plaque organisms, S. sanguis
(Gram-positive) and C. gingivalis (Gram-negative).
Materials and methods
Organisms and media
S. sanguis strain ATCC 10556 and C. gingivalis ATCC
33624 were maintained as stock cultures in tryptone (1.0%
w/v) yeast extract (0.5% w/v) broth (Oxoid, Basingstoke,
UK) plus glycerol (15% w/v) at –20°C. Samples taken
during continuous culture experiments were tested
regularly for purity by microscopic examination of Gramstained smears and plating of samples on to blood agar
(Oxoid). Plates were incubated either aerobically or in
an anaerobic cabinet (Don Whitley, Shipley, UK) in an
atmosphere of nitrogen, hydrogen and carbon dioxide
(80:10:10, by volume) at 37°C for 2 days. For continuous
culture studies, half-strength brain–heart infusion (BHI)
yeast extract plus glucose basal medium was used consisting of BHI (Difco, East Molesey, UK) (18.5 g/L), yeast
extract (5 g/L) and glucose (2 g/L). A stock solution of
glucose was steam-sterilized at 110°C for 30 min and added
aseptically to the rest of the medium which had previously
been sterilized by autoclaving. Triclosan was dissolved in
1 M NaOH which was then diluted (typically up to 5 L) and
neutralized to pH 7.5 using 1 M/0.1 M HCl. Stock solutions
of triclosan were sterilized by autoclaving. TMP stock solutions were sterilized by membrane filtration (Sartobran
capsule filter, 0.2 m porosity; Sartorius, Epsom, UK). Triclosan or TMP was aseptically added progressively to the
bulk medium, previously sterilized by autoclaving to give a
final volume of 20 L and a range of final concentrations
from 0 to 20 mg/L for triclosan and 0 to 50 mg/L ‘triclosan
equivalent’ concentrations for TMP (based on the knowledge that 70% of the weight of TMP is triclosan). In these
experiments, the stated concentrations are the amounts
added; no attempt was made to measure the bioavailable
concentrations of triclosan or TMP.
Chemicals
All chemicals except polypropylene glycol antifoam
(BDH, Poole, UK), triclosan and TMP (Procter & Gamble,
Cincinnati, OH, USA) were obtained from Sigma (Poole,
UK).
Continuous culture apparatus and conditions
S. sanguis was grown in a 1 L culture vessel with control
modules for temperature, pH, gas flow and stirrer rate
(Series 500, L.H. Engineering, Stoke Poges, UK). The
culture volume was maintained at 750 mL, temperature at
37°C ( 0.1°C) and the pH at 7.5 ( 0.1 unit) by the automatic addition of 2 M NaOH or 2 M H2SO4. Foaming was
controlled by the addition of antifoam when necessary. The
impeller speed was 600 rpm ( 10 rpm) and the vessel was
sparged with oxygen-free nitrogen plus carbon dioxide
(90:10, v/v) at a flow rate of 100 mL/min ( 5.0 mL/min).
The same conditions were employed for the growth of
C. gingivalis with the exception that the vessel was sparged
with a gas mixture of nitrogen, carbon dioxide and oxygen
(80:10:10 by volume).
660
Triclosan/TMP effects on S. sanguis and C. gingivalis
Determination of biomass and maximum specific
growth rates
These were determined as described previously.8 In determining the maximum specific growth rates, the dilution
rate was set at 1.30/h for each washout experiment.
determining the max and then running the chemostat at a
dilution rate to give 0.1 max was repeated for each triclosan
or TMP concentration used. Therefore, biomass and
hydrolytic enzyme activities were compared for each condition when organisms were growing at the same relative
growth rate (0.1 rel).
Culture fraction preparation
Statistical analysis
Samples taken from the chemostat were centrifuged at
3000g for 30 min to give two fractions: cells and supernatant. The enzyme activities measured in these two
fractions were denoted as ‘cell-bound’ and ‘extracellular
soluble’, respectively.
Linear regression analysis was used to determine the slope
of the curve for the washout data for calculation of max.
For hydrolytic enzyme production, eight determinations
were made for each fraction of each sample. From these
data the means and standard deviations were computed
and the differences between states analysed and shown to
be statistically significant using ANOVA.
Enzyme assays
Acid phosphatase (EC 3.1.3.2) activity was measured by a
method9 using the chromogenic substrate p-nitrophenyl
phosphate in 0.1 M 2-(N-morpholino)ethane sulphonic
acid (MES) buffer at pH 5.25 (for S. sanguis) and 0.1 M 3(N-morpholino)propane sulphonic acid (MOPS), pH 7.5
for C. gingivalis. Leucine aminopeptidase (EC 3.4.1.1)
activity was measured spectrophotometrically10 using
the synthetic chromogenic substrate, leucyl-p-nitroanilide,
which upon hydrolysis produces the coloured end-product
p-nitroaniline ( max 410 nm). Assays were carried out in
0.1 M MES at pH 6.5 (S. sanguis) or 0.1 M MOPS, pH 7.5
(C. gingivalis). Esterase activity (EC 3.1.1.1) (S. sanguis)
was measured in the same manner as phosphatase activity
except that the substrate used was p-nitrophenyl butyrate
in 0.1 M 3-(cyclohexylamino)-1-propane sulphonic acid
(CAPS) buffer at pH 9.0. Trypsin-like protease activity of
C. gingivaliswas measured10 using N- -benzoyl-L-argininep-nitroanilide which upon hydrolysis releases p-nitroaniline ( max
410 nm). Assays were carried out in
0.1 M MOPS, pH 7.5. For all enzyme activities, initial
experiments were carried out to determine the optimal
pH for activity and subsequent routine assay.
Results
S. sanguis
Maximum specific growth rates. The max values (Figure 1)
decreased with increasing concentrations of triclosan. At
the highest concentration of triclosan used, i.e. 20 mg/L,
steady-state growth could not be maintained even at low
dilution rates. Subsequently, the medium pump was
switched off to put the culture into batch culture mode. The
culture failed to regrow and was clearly inhibited by this
level of triclosan. Consequently, the values for max and
biomass were recorded as zero.
The max values first increased and then decreased with
increasing concentrations of TMP with the exception that
the concentration range was extended from 0 to 50 mg/L
before inhibition of growth was apparent. Thus, TMP was
less inhibitory against S. sanguis than triclosan was.
Culture biomass. The culture biomass (Figure 2) measured
at 0.1 rel remained approximately constant with increasing
concentrations of triclosan. For TMP, biomass levels
Units of enzyme activity
The unit of activity (U) for bacterial cells or supernatant
fractions derived from a known quantity of cells is calculated as mol end-product/h. Activities are expressed as
specific activity, in U/mg biomass.
Steady state
The maximum specific growth rate ( max) was determined
initially for controls lacking triclosan and TMP. Thereafter
the dilution rate was set to allow for a relative growth rate
of 0.1 max. A steady state was considered to have been
achieved after a minimum of six pot-volume changes had
occurred and variation between sample readings for Figure 1. Effects of triclosan ( ) and TMP ( ) on the maxienzymes and biomass were minimal. The procedure of first mum specific growth rates of S. sanguis.
661
J. Greenman et al.
Table I. Effects of triclosan on the proportions of cellbound (CB) and extracellular soluble (ES) hydrolytic
enzyme activities (expressed as CB/ES percent total
ratios) produced by S. sanguis
Figure 2. Effects of triclosan ( ) and TMP ( ) on S. sanguis
culture biomass at 0.1 rel.
[Triclosan]
(mg/L)
Acid
phosphatase
Leucine
aminopeptidase
Esterase
0
5
10
15
17.5
20
97.8/2.2
97.7/2.3
96.7/3.3
96.0/4.0
95.9/4.1
–
89.8/10.2
79.9/20.1
82.4/17.6
72.9/27.1
64.8/35.2
–
28.1/71.9
32.8/67.2
34.6/65.4
44.2/55.8
39.0/61.0
–
Table II. Effects of TMP on the proportions of cellbound (CB) and extracellular soluble (ES) hydrolytic
enzyme activities (expressed as CB/ES percent total
ratios) produced by S. sanguis
Figure 3. Effects of triclosan ( ) and TMP ( ) on S. sanguis
acid phosphatase specific activity at 0.1 rel.
remained approximately constant from 0 to 20 mg/L TMP.
At higher levels of TMP (30 and 40 mg/L) the biomass
levels dropped significantly compared with the control
condition.
Acid phosphatase. Acid phosphatase activity was found
mainly as cell-bound activity (Tables I and II) with <5% of
the total activity being detected in the supernatant at any
time. The levels of acid phosphatase activity (Figure 3)
decreased with increasing concentrations of triclosan
although the decrease was statistically significant only at
antimicrobial concentrations of 15 and 17.5 mg/L. The
proportion of extracellular soluble activity (Table II)
increased slightly with increases in the triclosan concentrations although at no time did it exceed 5% total activity.
For TMP, phosphatase activity remained approximately
constant with increasing concentrations of antimicrobial
until a concentration of 30 mg/L was reached. At this point
the enzyme activity increased above the levels in the
control. At higher concentrations of TMP (40 mg/L) phosphatase activity was again decreased. The proportion of
[TMP]
(mg/L)a
Acid
phosphatase
Leucine
aminopeptidase
Esterase
0
5
10
20
30
40
96.9/3.1
95.2/4.8
99.2/0.8
97.3/2.7
97.6/2.4
97.2/2.8
88.1/11.9
86.8/13.2
87.3/12.7
70.6/29.4
65.2/34.8
94.8/5.2
33.4/66.6
34.5/65.5
34.6/65.4
32.3/67.7
31.0/69.0
18.5/81.5
a
Triclosan equivalent.
soluble (extracellular soluble) activity remained low
(Table II).
Leucine aminopeptidase. Leucine aminopeptidase activity
was predominantly cell-bound (Tables I and II) but its pattern of production differed from that of acid phosphatase
since the levels found in the supernatant as extracellular
soluble activity were comparatively high (10–35% total
activity). Total activity (cell-bound plus extracellular soluble) increased as the triclosan concentration increased
(Figure 4) and the levels detected in the supernatant (extracellular soluble activity) increased as a proportion of the
cell-bound activity (Table II).
For TMP, levels of leucine aminopeptidase decreased
with an irregular trend as TMP concentrations increased
(Figure 4). The proportion of extracellular soluble activity
(Table II) remained approximately constant with increasing
concentrations of TMP to 20 mg/L and 30 mg/L where
it increased considerably. At a higher TMP concentration
(40 mg/L) the proportion of extracellular soluble aminopeptidase decreased to less than the control (zero TMP).
662
Triclosan/TMP effects on S. sanguis and C. gingivalis
Esterase. In contrast to acid phosphatase and leucine
aminopeptidase, esterase activity was detected mainly as
soluble (extracellular soluble) activity, typically 55–80% of
the total activity (Tables I and II). For example, in control
conditions, the proportion of extracellular soluble activity
was over twice the cell-bound levels. For both triclosan and
TMP the trends for total esterase activity (Figure 5) were
very similar to those shown for acid phosphatase activity.
Thus for triclosan, levels remained constant across
the range of triclosan concentrations used. For TMP,
esterase activity remained approximately constant for TMP
concentrations of 0 to 20 mg/L but then increased at a
TMP concentration of 30 mg/L.
When expressed as a proportion of total activity (Table
II) the extracellular soluble esterase activity was seen to
remain approximately constant with increasing concentrations of triclosan until a triclosan concentration of 15 mg/L
was reached, at which the proportion decreased. At a
higher triclosan concentration (17.5 mg/L) the extracellular
soluble proportion increased again. For TMP, the proportion of extracellular esterase remained approximately
constant until the TMP concentration reached 40 mg/L
where the proportion was significantly reduced.
Culture biomass. The steady-state culture biomass levels
measured at 0.1 rel for C. gingivalis (Figure 7) increased
with increases in triclosan concentration from 0 to 7.5
mg/L. At the higher triclosan concentration of 10 mg/L the
biomass levels decreased significantly to 30% of those
measured for the control (zero triclosan) condition. The
lethal concentration of triclosan in the chemostat was shown
to be between 10 and 12.5 mg/L. Increasing levels of TMP
Figure 4. Effects of triclosan ( ) and TMP ( ) on S. sanguis
leucine aminopeptidase specific activity at 0.1 rel.
Figure 6. Effects of triclosan ( ) and TMP ( ) on the maximum specific growth rates of C. gingivalis.
Figure 5. Effects of triclosan ( ) and TMP ( ) on S. sanguis
esterase specific activity at 0.1 rel.
Figure 7. Effects of triclosan ( ) and TMP ( ) on C. gingivalis
culture biomass at 0.1 rel.
C. gingivalis
Maximum specific growth rates. The max values (Figure 6)
showed marked reductions with increases in the test concentrations of triclosan and TMP. However, C. gingivalis
appeared to be more susceptible than S. sanguis, to the two
test compounds, and was more susceptible to TMP (lethal
level 7.5 mg/L) than to triclosan (lethal level 12.5 mg/L).
663
J. Greenman et al.
Table III. Effects of triclosan on the proportions of cellbound (CB) and extracellular soluble (ES) hydrolytic
enzyme activities (expressed as CB/ES percent total
ratios) produced by C. gingivalis
[Triclosan]
Leucine
Trypsin-like
(mg/L)
Phosphatase aminopeptidase protease
0
2.5
5.0
7.5
10
12.5
84.7/15.3
93.6/6.4
90.4/9.6
84.9/15.1
34.9/65.1
–
92.7/7.3
95.0/5.0
96.1/3.9
94.1/5.9
63.5/36.5
–
87.4/12.6
89.6/10.4
82.6/17.4
79.5/20.5
53.5/46.5
–
Figure 8. Effects of triclosan ( ) and TMP ( ) on C. gingivalis
neutral phosphatase specific activity at 0.1 rel.
Table IV. Effects of TMP on the proportions of cellbound (CB) and extracellular soluble (ES) hydrolytic
enzyme activities (expressed as CB/ES percent total
ratios) produced by C. gingivalis
[TMP]
(mg/L)a
Phosphatase
Leucine
Trypsin-like
aminopeptidase protease
0
1.25
2.5
5.0
7.5
84.7/15.3
85.5/14.5
90.3/9.7
61.9/38.1
–
92.7/7.3
93.5/6.5
94.7/5.3
61.1/38.9
–
87.4/12.6
84.3/15.7
83.0/17.0
54.2/45.8
–
a
Triclosan equivalent.
(from 0 to 5.0 mg/L) produced decreasing values for biomass. The lethal concentration was between 5 and 7.5 mg/L,
a value considerably less than that obtained for triclosan.
Phosphatase. Phosphatase activity was mainly detected as
cell-bound activity (Tables III and IV). Activity (Figure 8)
remained approximately constant at 0–2.5 mg/L triclosan
but then decreased with increasing concentrations of triclosan. Expressed as a proportion of total activity (Table
III), extracellular soluble activity varied between 6 and 15%
at triclosan concentrations from 0 to 7.5 mg/L. However,
the extracellular soluble proportion increased to 65% of
the control value at the highest concentration of triclosan
(10 mg/L).
Phosphatase activity appeared to increase with increasing concentrations of TMP (Figure 8), but then returned to
values below that of the control when grown at a TMP
concentration of 5.0 mg/L. Extracellular soluble activity,
expressed as a proportion of total activity (Table IV),
decreased slightly, but then increased significantly as the
TMP concentration increased to 5.0 mg/L. At this point
nearly 40% of the total activity was due to extracellular
soluble activity.
Figure 9. Effects of triclosan ( ) and TMP ( ) on C. gingivalis
leucine aminopeptidase specific activity at 0.1 rel.
Leucine aminopeptidase. For leucine aminopeptidase,
activity was mainly cell-bound. Total culture activities
(Figure 9) were reduced with increases in the concentrations of triclosan from 0 to 7.5 mg/L. At the higher triclosan
concentration (10 mg/L), activity increased, reaching those
found in the control (zero triclosan) condition. Expressed
as a proportion of total activity (Table III), extracellular
soluble activity remained low (<6%) until a triclosan concentration of 10 mg/L was reached. At this point the extracellular soluble proportion showed a marked increase,
representing 36.5% of total culture activity.
For TMP, the pattern of leucine aminopeptidase (Figure
9) was similar to that shown for triclosan, in that the total
culture activity at first decreased (at TMP concentrations
of 0–2.5 mg/L) but then increased again when the TMP
concentration reached 5 mg/L. The decrease followed by
an increase was most marked for extracellular soluble
activity. At 5 mg/L TMP the release of aminopeptidase
from cells was clearly apparent (Table IV), with extracellular soluble activity reaching 40% the value of total culture
activity.
664
Triclosan/TMP effects on S. sanguis and C. gingivalis
Figure 10. Effects of triclosan ( ) and TMP ( ) on C. gingivalis
trypsin-like protease (TLPase) specific activity at 0.1 rel.
Trypsin-like protease. Trypsin-like protease activity was
reduced with increasing concentrations of triclosan from 0
to 7.5 mg/L (Figure 10). At 10 mg/L triclosan, activity
increased again and a significant proportion of activity was
now extracellular soluble (Table III).
Compared with control conditions, trypsin-like protease
activity decreased with increasing concentrations of TMP
(Figure 10). As seen with triclosan, in experiments using
the highest concentration of test agent that could still
support chemostat growth (10 mg/L for triclosan; 5 mg/L
for TMP), a significant proportion of trypsin-like enzyme
was released from the cell as extracellular soluble activity
(Table IV).
Discussion
In this study, the assumption was made that the cellular and
molecular mechanisms of antimicrobial action are similar
at both lethal and sub-lethal concentrations of antimicrobial agent. Although this assumption is reasonable it
may not always be true. Moreover, indirect or secondary
effects of antimicrobial exposure may occur in cells and
may merely indicate that the cell is undergoing general
metabolic stress.
The initial tube MICs of triclosan against S. sanguis and
C. gingivalis, using an inoculum which gave a final concentration of approximately 106 cells/mL, has been shown to
be 5 mg/L and 2.5 mg/L respectively.4 In the chemostat, the
concentration of cells present when the antimicrobial is
introduced into the culture is very much higher (by a factor
of approximately 100). Since the MICs are likely to be at
least partly dependent on the ratio of target sites (related to
cell concentration) to molecules of antimicrobial present
(i.e. concentration of triclosan), it is not surprising that the
inhibitory concentrations determined in the chemostat (20
and 12.5 mg/Lfor S. sanguis and C. gingivalis respectively)
were higher than their corresponding tube MICs. It is more
difficult to explain why the inhibitory level determined in
the chemostat is not even higher (e.g. 100-fold more than
the MICs, reflecting the 100-fold increase in cells). One
possible reason for this is that cells are more susceptible to
antimicrobial when they are actively dividing (which is
always the case in continuous culture) and this would make
individual cells more susceptible to antimicrobial.
Both S. sanguis and C. gingivalis are clearly affected by
sub-lethal concentrations of triclosan and TMP. What is
surprising is that the maximum specific growth rate of S.
sanguis at low concentrations of triclosan was higher than
the control (zero agent) condition. This suggests that at low
concentrations, triclosan may bind to the membrane and
change its properties, enhancing cell transport or membrane fluidity. For both S. sanguis and C. gingivalis, culture
biomass (measured at 0.1 rel) increases in some cases with
increasing concentrations of agents. This finding is similar
to that of Minhas & Greenman6 who tested sub-lethal
concentrations of chlorhexidine against the Gram-negative
anaerobe, Porphyromonas gingivalis. In that study, the
biomass levels also increased slightly at sub-lethal concentrations of antimicrobial. However, in all cases the
maximum specific growth rates and biomass decreased at
concentrations of agents approaching the lethal level,
suggesting that cells are being subjected to metabolic stress.
When grown in pure culture, the phosphatase activities
of S. sanguis and C. gingivalis will be of central importance
in converting the relatively inactive TMP into its active
form, triclosan. This will occur at the culture pH, 7.5, which
is close to the optimum for C. gingivalis4 but not so for
S. sanguis. Therefore, conversion of TMP into triclosan is
likely to be more efficient for C. gingivalis than for S. san guis. Furthermore, the synthesis, expression and activity of
S. sanguis phosphatase may itself be sensitive to the effects
of triclosan since lower levels of activity were measured
when grown in the presence of 15–17.5 mg/L triclosan. All
these factors may help explain why (compared with triclosan) TMP is much more effective against C. gingivalis
and much less effective against S. sanguis.
It is difficult to explain why, for C. gingivalis, TMP
should be more inhibitory than triclosan. It is possible that
the highly soluble TMP can diffuse into the periplasmic
space of C. gingivalis where it may be hydrolysed by
membrane-bound phosphatases into triclosan which
remains relatively localized or trapped, possibly adjacent
to triclosan-target sites. Triclosan, however, because of
its lipophilicity/solubility cannot diffuse as easily to the
triclosan-target sites, and hence is less effective. This may
not occur in Gram-positive organisms such as S. sanguis.
The reduction of maximum specific growth rate of cells
with increasing concentrations of triclosan and TMP indicates that general metabolic stress is occurring which may
in turn affect the level of hydrolytic enzyme expression. In
some cases, enzyme expression increased at low concentrations of antimicrobial agents. This may reflect a change in
the activity or stability of the enzymes located on the
665
J. Greenman et al.
conditioned membrane. It is not known if mechanisms exist
within microorganisms whereby release of an enzyme by
the cell can improve or promote the efficiency of enzyme
synthesis or translocation in a compensatory system to
‘replace’ lost enzyme. For S. sanguis, TMP at low concentrations could act as an inducer as well as a substrate of
phosphatase activity.
Increases in the proportions of extracellular soluble
enzymes ‘released’ by the cell indicate a more specific
perturbation of membrane functions. The observation that
the ratio of extracellular soluble to cell-bound hydrolytic
enzyme activities increased in most cases, with increases in
triclosan/TMP concentration indicates that changes are
occurring in the cell membrane structures that are associated with the sites of location of these enzymes. This is not
incompatible with what is known about the action of
triclosan. It is reported to bind to the lipid portion of the
bacterial cytoplasmic membrane and to interfere with
transport mechanisms.3 The different membrane structure
of Gram-positive and Gram-negative cell walls may also
explain why C. gingivalis (which is Gram-negative) releases
more hydrolytic enzymes when subjected to increasing
concentrations of antimicrobial agent than does S. sanguis
(which is Gram-positive). Both show some release of
hydrolytic enzymes suggesting that membrane perturbations are occurring, but this is much more marked for
C. gingivalis. For this species, the hydrolytic enzymes are
probably located within the periplasmic space, typical of
Gram-negative organisms.11 Damage to the outer membrane and associated release of enzymes may occur at an
antimicrobial concentration lower than that required to
damage the inner cytoplasmic membrane. This cannot
occur with S. sanguis since enzyme release is integral with
damage to the single membrane target, the cytoplasmic
membrane. Thus, the point between enzyme release and
lethality may be much closer for this species.
The results of this investigation confirm the hypothesis
that antimicrobial agents affect microbial physiology at
sub-lethal concentrations. For periodontopathic bacteria
such as Capnocytophaga spp., where virulence is thought to
be attributable mainly to expression of hydrolytic enzymes,
the results of this study strongly suggest that antimicrobial
agents may be active in reducing the virulence of organisms
at sub-lethal concentrations.
References
1. Regös, J. & Hitz, H. R. (1974). Investigation of mode of action
of triclosan, a broad spectrum antimicrobial agent. Zentralblatt
für Bakteriologie, Parasitenkunde, Infektionskrankheiten und
Hygiene IA (Medizinsche, Mikrobiologie und Parasitologie) 226,
390–401.
2. Vischer, W. A. & Regös, J. (1974). Antimicrobial spectrum of
triclosan, a broad-spectrum antimicrobial agent for topical application. Zentralblatt für Bakteriologie, Parasitenkunde, Infektion skrankheiten und Hygiene IA (Medizinsche, Mikrobiologie und
Parasitologie) 226, 376–89.
3. Meincke, B. E., Krantz, R. G. & Lynch, D. L. (1980). Effect of
irgasan on bacterial growth and its adsorption into the cell wall.
Microbios 28, 133–47.
4. Greenman, J. & Nelson, D. G. A. (1996). Hydrolysis of triclosan
monophosphate by dental plaque and selected species of oral
micro-organisms. Journal of Dental Research 75, 1578–84.
5. Greenman, J. (1984). Effects of sub-lethal concentrations of the
antimicrobial agent propylene phenoxetol on the growth and extracellular enzymes of Propionibacterium acnes. Microbios 39,
101–8.
6. Minhas, T. & Greenman, J. (1989). The effects of chlorhexidine
on the maximum specific growth rate, biomass and hydrolytic
enzyme production of Bacteroides gingivalis grown in continuous
culture. Journal of Applied Bacteriology 67, 309–16.
7. Greenman, J. & Minhas, T. (1990). The effects of tetradecyl-4ethyl-pyridinium chloride on the maximum specific growth rate,
biomass and hydrolytic enzyme production of Bacteroides
gingivalis in continuous culture. Journal of Antimicrobial Chemo therapy 26, 649–57.
8. Minhas, T. & Greenman, J. (1989). Production of cell-bound
and vesicle-associated trypsin-like protease, alkaline phosphatase
and N-acetyl- -glucosaminidase by Bacteroides gingivalis W50.
Journal of General Microbiology 135, 557–64.
9. Holland, K. T., Greenman, J. & Cunliffe, W. J. (1979). Growth of
cutaneous propionibacteria on synthetic medium; growth yields
and exoenzyme production. Journal of Applied Bacteriology 47,
383–94.
10. Spratt, D. A., Greenman, J. & Schaffer, A. G. (1996). Capno cytophaga gingivalis: effects of glucose concentration on growth
and hydrolytic enzyme production. Microbiology 142, 2161–4.
11. Lopes, J., Gottfried, S. & Rothfield, L. (1972). Leakage of
periplasmic enzymes by mutants of Escherichia coli and
Salmonella typhimurium: isolation of ‘periplasmic leaky’ mutants.
Journal of Bacteriology 109, 520–5.
Acknowledgement
This investigation was supported by a grant from Procter
and Gamble, Cincinnati, OH, USA.
Received 7 January 1997; returned 17 February 1997; revised 2
April 1997; accepted 16 June 1997
666