JAC
Journal of Antimicrobial Chemotherapy (1997) 40, 77–83
A flow cytometric study of antibiotic-induced damage and evaluation
as a rapid antibiotic susceptibility test for methicillin-resistant
Staphylococcus aureus
M. T. E. Sullera* J. M. Starkb and D. Lloyda
a
School of Pure and Applied Biology, University of Wales College of Cardiff, PO Box 915, Cardiff CF1 3TL;
b
Department of Medical Microbiology, University of Wales College of Medicine, Heath Park,
Cardiff CF4 4XN, UK
Flow cytometry using the anionic membrane potential-sensitive fluorescent probe, bis-(1,3dibutylbarbituric acid) trimethine oxonol (DiBAC4(3)), enabled assessment of antibiotic-induced membrane perturbation in five clinical isolates of methicillin-resistant
Staphylococcus aureus (MRSA) and two antibiotic-sensitive reference strains, NCTC 6571
and 8325-4, after establishment of steady-state growth in liquid cultures inoculated from
single colonies. Flow cytometric indications of the enhanced DiBAC4(3) uptake after treatment with vancomycin at 0.1, 1, 4 and 10 x MIC showed excellent comparison with viability
losses quantified as cfu on solid agar in MRSA isolate QC. The antibiotic susceptibility
patterns to benzylpenicillin, methicillin and vancomycin for all isolates used in this study
could be determined in 2–4 h from an overnight plate culture. This technique thus provides a
rapid and reproducible antibiotic sensitivity test which may be applicable in routine clinical
practice.
of rapid antibiotic susceptibility tests would allow specific
and cheaper narrow-spectrum antibiotics to be used, resulting in improved patient care and cost-effectiveness.
Flow cytometry has gained increasing importance in
many aspects of biomedical science.7,8 It allows the rapid
analysis of properties of individual cells in non-synchronous
cultures with excellent statistics. The potential of flow
cytometry to provide a rapid test of bacterial antibiotic
susceptibility was explored by Steen et al.,9 who demonstrated the effect of antibiotics on Escherichia coli with the
fluorescent DNA probe mithramycin. Subsequently, flow
cytometry was used to assess antibiotic-induced membrane
damage10 and antibiotic-induced inhibition of E. coli within
minutes.11 Recent flow cytometric susceptibility tests have
employed membrane potential-sensitive fluorescent
probes to assess membrane damage and depolarization.12–20
Oxonols are lipophilic anionic dyes that enter cells with
depolarized plasma membranes and bind to lipid-rich intracellular components; fluorescence is enhanced upon
accumulation.18 They have been used recently to determine
the viability of Trichomonas vaginalis21 and to evaluate the
effects of antibiotics on Candida albicans22 and some
pathogenic bacteria.13
Introduction
Resistance of Staphylococcus aureus to antibiotics is a
serious and growing problem in terms of human suffering
and economics. 1,2 In 1941, all known strains were susceptible to benzylpenicillin (penicillin G). However, by 1992,
95% of clinical isolates were resistant to penicillin G and
ampicillin because of the production of a -lactamase.3
Methicillin, a semi-synthetic penicillin, was introduced to
clinical practice in 1959, but by 1961 methicillin-resistant
strains had been isolated.4 In 1980, the extent of the
problem became apparent when methicillin-resistant
S. aureus (MRSA) resistant to aminoglycosides such as
gentamicin was first described. Methicillin resistance in
S. aureus results from a chromosomally located mecA gene
that encodes a new penicillin-binding protein (PBP-2 )
with a low affinity for -lactam antibiotics.5 Vancomycin is
the antibiotic of choice against these strains and, to date,
resistance has not been reported.
Traditional methods of antibiotic susceptibility testing
require at least 48 h to provide the physician with a result.
During this period, the patient may be treated with inappropriate broad-spectrum antibiotics.6 The development
*Tel:
44-1222-874000 ext. 6350; Fax:
77
© 1997 The British Society for Antimicrobial Chemotherapy
44-1222-874305.
M. T. E. Suller et al.
Flow cytometry
The present work illustrates the use of the anionic
membrane potential-sensitive fluorescent probe, bis-(1,3dibutylbarbituric acid) trimethine oxonol (DiBAC4(3) ) to
assess by flow cytometry the antibiotic-induced perturbation of the plasma membrane of clinical isolates of MRSA
and thereby provide a rapid and reproducible antibiotic
susceptibility test.
Flow cytometry was performed on a Skatron Argus 100
high pressure mercury arc lamp based, dual parameter flow
cytometer (Skatron Instruments Ltd, Newmarket, UK).
The sheath fluid pressure was set to 1 KPa/cm2 and the
sample flow rate to 5 L/min. The fluorescence excitation
and emission maximum wavelengths for DiBAC4(3) are
493 nm and 516 nm, respectively. Fluorescence was
detected using a FITC filter block (Skatron) which provides an excitation wavelength of 470–490 nm, a band stop
of 510 nm and an emission of 520–550 nm. Data acquisition
of fluorescence was gated by forward angle light scatter,
and detected by linear amplification. The forward angle
light scatter detector, an indication of cell size, was used
with logarithmic amplification. Samples were allowed to
run for 1 min, before data acquisition from 5000 organisms.
The performance of the flow cytometer was monitored
daily using 1 m fluorescent beads (Park Scientific Ltd,
Northampton, UK).
Materials and methods
Bacterial strains and growth conditions
Five clinical isolates of MRSA (QC, NCTC 11353, Z60699,
U472137 and U472011) were provided by Mr Alan Paull of
the Department of Medical Microbiology, University of
Wales College of Medicine. The reference strains used
were S. aureus NCTC 6571 and S. aureus 8325-4, both of
which are sensitive to penicillin G, methicillin and vancomycin. All isolates were maintained on nutrient agar
(Oxoid, Basingstoke, UK). For flow cytometric analysis,
bacteria were grown to early exponential growth phase
(optical density of 0.2) in filtered (0.2 m) nutrient broth
(Oxoid) in a reciprocating water bath at 37°C and at 100
strokes per minute.
Flow cytometric antibiotic susceptibility test
Flow cytometric susceptibility tests were performed as
outlined by Mason et al.13 After being subcultured from a
single colony, each isolate was grown to an optical density
of 0.2 (600 nm) which indicates early exponential growth
phase. The antibiotic was then added and 1 mL samples
removed at timed intervals and stained with DiBAC4(3)
before flow cytometric analysis.
Antibiotics and MIC testing
The antibiotics used were penicillin G, methicillin and
vancomycin (all from Sigma, Poole, UK). For flow cytometric susceptibility testing, concentrations at the breakpoints recommended in clinical practice were used (0.12
mg/L, 4 mg/L and 4 mg/L for penicillin G, methicillin and
vancomycin, respectively). Independent susceptibility tests
were performed using the broth dilution technique where
the bacteria are incubated in nutrient broth with a series of
antibiotic concentrations.
Viable counts
Counts of cfu were performed as outlined by Miles &
Misra23 after appropriate dilution in HEPES buffer
(Sigma).
Results
The antibiotic-sensitive reference strains and all the
MRSA clinical isolates showed increased fluorescence
following treatment with heat or gramicidin. This indicated
complete membrane depolarization, allowing entry of the
dye, and underlined the possible use of this procedure to
assess destruction of the plasma membrane dye exclusion
barrier in the different isolates of S. aureus. The effect of
heat and gramicidin treatment on MRSA QC (a clinical
isolate) is shown in Figure 1, and was typical of all isolates
tested. The dual parameter histogram gives a threedimensional perspective of the population based on the
light scatter and fluorescence from individual cells, the
vertical axis representing the relative number of organisms.
The effect of methicillin on S. aureus NCTC 6571 is
shown in Figure 2a. A significant increase in fluorescence
was detected after 30 min incubation with the antibiotic.
No isolates of MRSA showed such a response at the corresponding time intervals, indicating that the methicillin was
Staining with DiBAC4(3)
A stock solution of DiBAC4(3) (Cambridge Biosciences,
Cambridge, UK) was prepared in ethanol at a concentration of 1 g/L and stored at 20°C. The dye was added
directly to the broth culture to give a final concentration of
1 mg/L. Incubation was for 2 min at room temperature
before flow cytometric analysis.
Gramicidin S and heat treatment
As a control experiment to assess the effectiveness of the
dye to differentiate between viable and dead cells, all isolates were heat-treated (at 70°C for 30 min) and incubated
in the presence of the ionophore gramicidin S (20 mg/L for
15 min). This results in the depolarization of the bacterial
plasma membrane allowing entry of the dye into the cell
and intense intracellular fluorescence.
78
Flow-cytometric study of MRSA
of the sensitivity of that population to a particular antibiotic. Penicillin and methicillin caused an increased
fluorescence in the two reference strains, but not the
MRSA. Vancomycin, on the other hand, did give increased
fluorescence emission in all the MRSA strains, and this
demonstrates the susceptibility of MRSA to this antibiotic.
Vancomycin-induced damage to MRSA isolate QC in
terms of increased fluorescence and decreased viability
assessed by cfu is shown in Figure 3. If treatment with the
antibiotic results in membrane potential disruption and
loss of viability, an increase in fluorescent intensity will be
observed. The antibiotic at 4
MIC (the ‘breakpoint
value’) and 10
MIC resulted in median fluorescence
values of 84 and 91, respectively, after 4 h incubation and
this was paralleled by the loss of cell viability based on cfu
counts. Incubation at the MIC value resulted in a smaller
increase in fluorescence after 4 h at a median channel of 45.
No fluorescence increase was seen at 0.1 MIC and no loss
of viability was recorded from plate counts.
Discussion
These results demonstrate the successful use of flow cytometric measurement of DiBAC 4(3) fluorescence to detect
changes in the transmembrane electrochemical potential
produced by the action of antibiotics on S. aureus. This
provided a rapid and reproducible antibiotic susceptibility
test that gave a result within 2–4 h of subculture from an
overnight plate culture. The rate-determining factor in this
procedure arose from the period required to establish
‘steady state’ exponential growth. This compared
favourably with conventional susceptibility tests that
require a second incubation overnight. However, a flow
cytometric assay of methicillin resistance would not distinguish between mecA-containing isolates and others that
show decreased susceptibility to methicillin. This information, often required by the clinical laboratory, may be
obtained using PCR.
MICs derived from conventional tests are open to correction; they may exhibit poor correlation with actual doseresponses24 and frequently do not truly represent the
‘minimum’ inhibitory concentration, but rather an intermediate value depending on the dilution series used. Also,
conventional tests do not distinguish between concentration- and time-dependent antibiotic activities; this ambiguity may have important implications for dosage
regimes.25 Conventional tests only express an average
susceptibility value for the whole bacterial population, but
this population may be heterogeneous in terms of cell
cycle, age, growth rates and metabolism.26,27 Mutations
may arise during growth and the frequency is increased on
exposure to selective pressures (e.g. the presence of an
antibiotic). On solid agar, a colony will show heterogeneity
often over very small distances. As flow cytometry can be
used to analyse individual cells in a population, it can pro-
Figure 1. Dual-parameter histograms of forward angle light
scatter (FALS) against DiBAC4(3) fluorescence for (a) viable,
(b) heat-treated and (c) gramicidin-treated cells of MRSA QC
(a clinical isolate).
not causing detectable perturbation to the plasma membrane. Similarly, Figure 2b shows the effect of vancomycin
on MRSA NCTC 11353 (clinical isolate) where susceptibility could again be detected after only 30 min.
The effects of penicillin G, methicillin or vancomycin on
all strains used are summarized in the Table. Fluorescence
is measured on an arbitrary scale of 0–250 each of which
corresponds to a channel number; thus the median fluorescent intensity (the channel number above and below which
50% of the distribution can be found)10 gives an indication
79
M. T. E. Suller et al.
80
Flow-cytometric study of MRSA
Table. MICs for clinical isolates of MRSA and antibiotic-sensitive reference strains of S. aureus; also shown are
median channel numbers for fluorescence emission after flow cytometric analysis of all isolates after 120 min
incubation with penicillin G (0.12 mg/L), methicillin (4 mg/L) or vancomycin (4 mg/L). Typical results are given, but
replicate experiments showed these results to be reproducible.
Isolate
NCTC 6571
NCTC 8325-4
MRSA QC
MRSA NCTC 11353
MRSA Z60699
MRSA U472137
MRSA U472011
MIC (mg/L)
penicillin G methicillin vancomycin
0.031
0.062
8
8
8
8
8
1
1
8
8
8
8
8
1
1
1
1
1
1
1
Figure 3. The effect of vancomycin at various concentrations on
S. aureus isolate QC (a) fluorescence median channel number
and (b) viable count. The concentrations used were ( ) 0.1
mg/L, ( ) 1 mg/L (the MIC value), ( ) 4 mg/L and ( ) 10 mg/L.
vide essential information on the heterogeneity of the population in terms of susceptibility to antibiotics and biocides,
etc. Thus, as individual cells are damaged there will be an
overall increase in fluorescence of DiBAC4(3) the extent of
which is time- and concentration-dependent. However, the
81
Median fluorescence intensity (channel
number) after treatment with
penicillin G
methicillin
vancomycin
45
51
13
11
14
13
11
88
87
12
10
12
13
15
81
86
94
98
91
75
88
method requires further evaluation against a panel of
MRSA expressing different levels of methicillin resistance
to ensure that it will correctly classify as resistant those isolates that have MICs of 8 and 16 mg/L.
The action of antibiotics that do not act directly on the
cytoplasmic membrane will also result in membrane
perturbation which may be measured by a membrane
potential-sensitive dye. This was confirmed by Mason et
al.13 who used DiBAC4(3) to analyse membrane damage
caused by gentamicin, a protein synthesis inhibitor, and
ciprofloxacin, a DNA gyrase inhibitor. Bacteriostatic
antibiotics are known to leave organisms intact, but at the
same time cause some membrane disturbance which may
be measurable by flow cytometry. Antibiotics at subinhibitory concentrations have also been shown to affect
bacteria28,29 although, in the present study, vancomycin at a
concentration of 0.1 MIC did not affect MRSA as judged
by flow cytometry and viable counts.
DiBAC4(3) has a number of advantages over other
membrane potential-sensitive probes. It can be added
directly to the liquid culture with no requirement for pretreatment. Rhodamine 123, a cationic membrane potential
dye, requires the organism to be washed, filtered and made
permeable by EDTA treatment in the case of Gramnegative bacteria.30,31 These pretreatments may cause additional perturbation to bacteria and interfere with the actual
antibiotic-induced damage. Rhodamine 123 and other
membrane potential dyes, such as those of the carbocyanine family, have proved unpredictable in the assessment of bacterial viability.32 DiBAC4(3), however, has
been used to assess bacterial viability of both Grampositive and Gram-negative bacteria, an essential property
in the development of a routine antibiotic susceptibility
test. Furthermore, the oxonols are anionic dyes and therefore can only enter a cell when the membrane has become
depolarized. Thus they are non-toxic to the organism.
Cationic dyes, on the other hand, are highly toxic.33 It could
be argued that the oxonol dye may cause further damage to
M. T. E. Suller et al.
a depolarized cell upon intracellular binding and alter the
effect of the antibiotic. However, Pore25 has proposed that
the oxonol-induced change may usefully amplify the
fluorescent response proportionally.
The more efficient use of existing antibiotics to delay the
emergence of bacteria such as vancomycin-resistant
S. aureus while more anti-staphylococcal drugs are developed is a paramount requirement for new diagnostic
technologies. With the increasing threat posed by MRSA
and other bacterial species, the use of a flow cytometer
provides a simple method for the rapid evaluation of antibiotic susceptibility patterns of clinical isolates.34 Even
shorter periods than those reported here can provide
highly significant evidence if perturbation occurs following
antibiotic exposure.
13. Mason, D. J., Allman, R., Stark, J. M. & Lloyd, D. (1994).
Rapid estimation of bacterial antibiotic susceptibility with flow
cytometry. Journal of Microscopy 176, 8–16.
14. Mason, D. J., Lopez-Amoros, R., Allman, R., Stark, J. M. &
Lloyd, D. (1995). The ability of membrane potential dyes and
calcafluor white to distinguish between viable and non-viable
bacteria. Journal of Applied Bacteriology 78, 309–15.
15. Jepras, R. I., Carter, J., Pearson, S. C., Paul, F. E. & Wilkinson, M. J. (1995). Development of a robust flow cytometric assay
for determining numbers of viable bacteria. Applied and Environ mental Microbiology 61, 2696–701.
16. Sims, P. J., Waggoner, A. S., Wang, C. H. & Hoffman, J. F.
(1974). Studies on the mechanism by which cyanine dyes
measure membrane potential in red blood cells and phosphatidylcholine vesicles. Biochemistry 13, 3315–30.
17. Zaritsky, A., Kihara, M. & Macnab, R. M. (1981). Measurement of membrane potential in Bacillus subtilis: a comparison of
lipophilic cations, rubidium ion, and a cyanine dye as probes.
Journal of Membrane Biology 63, 215–31.
Acknowledgements
This work was funded by the Welsh Scheme for the
Development of Health and Social Research (Welsh
Office). The authors would like to thank Mr Alan Paull for
providing the bacterial strains.
18. Epps, D. E., Wolfe, M. L. & Groppi, V. (1994). Characterization of the steady-state and dynamic fluorescence properties of the
potential-sensitive dye bis-(1,3-dibutylbarbituric acid) trimethine
oxonol (DiBAC4(3) ) in model systems and cells. Chemistry and
Physics of Lipids 69, 137–50.
19. Oyama, Y., Chikahisa, L., Tomiyoshi, F. & Hayashi, H. (1991).
Cytotoxic action of triphenyltin on mouse thymocytes: a flowcytometric study using fluorescent dyes for membrane potential
and intracellular Ca2 . Japanese Journal of Pharmocology 57,
419–24.
References
1. Cohen, M. L. (1992). Epidemiology of drug resistance: implications for a post-antimicrobial era. Science 257, 1050–5.
2. Kunin, C. M. (1993). Resistance to antimicrobial drugs—a
worldwide calamity. Annals of Internal Medicine 118, 557–61.
20. Wilson, H. A. & Chused, T. M. (1985). Lymphocyte membrane
potential and Ca2 -sensitive potassium channels described by
oxonol fluorescence measurements. Journal of Cellular Physiology
125, 72–81.
3. Neu, H. C. (1992). The crisis in antibiotic resistance. Science
257, 1064–73.
4. Brumfitt, W. & Hamilton-Miller, J. (1989). Methicillin-resistant
Staphylococcus aureus. New England Journal of Medicine 320,
1188–96.
21. Humphreys, M. J., Allman, R. & Lloyd, D. (1994). Determination of the viability of Trichomonas vaginalis using flow cytometry.
Cytometry 15, 343–8.
5. Kayser, F. H. (1995). Methicillin and glycopeptide resistance in
staphylococci: a new threat? Current Opinion in Infectious
Diseases 8, Suppl. 1, S7–11.
22. Carter, E. A., Paul, F. E. & Hunter, P. A. (1993). Cytometric
evaluation of antifungal agents. In Flow Cytometry in Microbiology
(Lloyd, D., Ed.), pp. 111–20. Springer-Verlag, London.
6. Granato, P. A. (1993). The impact of same-day tests versus
traditional overnight testing. Diagnostic Microbiology and Infectious
Disease 16, 237–43.
23. Miles, A. A., Misra, S. S. & Irwin, J. O. (1938). The estimation
of the bactericidal power of the blood. Journal of Hygiene 38,
732–49.
7. Boye, E., Steen, H. B. & Skarstad, K. (1983). Flow cytometry of
bacteria: a promising tool in experimental and clinical microbiology.
Journal of General Microbiology 129, 973–80.
24. Wexler, H. M. (1993). Susceptibility testing of anaerobic bacteria—the state of the art. Clinical Infectious Diseases 16, Suppl. 4,
S328–33.
8. Allman, R. (1993). Flow cytometry: principles and applications.
Proceedings of the Royal Microscopical Society 28, 45–54.
25. Pore, R. S. (1994). Antibiotic susceptibility testing by flow
cytometry. Journal of Antimicrobial Chemotherapy 34, 613–27.
9. Steen, H. B., Boye, E., Skarstad, K., Bloom, B., Godal, T. &
Mustafa, S. (1982). Applications of flow cytometry on bacteria: cell
cycle kinetics, drug effects, and quantitation of antibody binding.
Cytometry 2, 249–57.
26. Lloyd, D. (1993). Flow cytometry: a technique waiting for
microbiologists. In Flow Cytometry in Microbiology (Lloyd, D., Ed.),
pp. 1–9. Springer-Verlag, London.
10. Shapiro, H. M. (1988). Practical Flow Cytometry, 2nd edn.
Alan, R. Liss, New York, NY.
11. Boye, E. & Lobner-Olesen, A. (1991). Bacterial growth control
studied by flow cytometry. Research in Microbiology 142, 131–5.
27. Gant, V. A., Warnes, G., Phillips, I. & Savidge, G. F. (1993).
The application of flow cytometry to the study of bacterial
responses to antibiotics. Journal of Medical Microbiology 39,
147–54.
12. Ordonez, J. V. & Wehman, N. M. (1993). Rapid flow cytometric antibiotic susceptibility assay for Staphylococcus aureus.
Cytometry 14, 811–8.
28. Tornqvist, I. O., Holm, S. E. & Cars, O. (1990). Pharmacodynamic effects of subinhibitory antibiotic concentrations. Scandi navian Journal of Infectious Diseases, Suppl. 74, 94–101.
82
Flow-cytometric study of MRSA
29. Greenwood, D. (1981). In vitro veritas? Antimicrobial susceptibility tests and their clinical relevance. Journal of Infectious
Diseases 144, 380–5.
33. Diaper, J. P., Tither, K. & Edwards, C. (1992). Rapid assessment of bacterial viability by flow cytometry. Applied Microbiology
and Biotechnology 38, 268–72.
30. Matsuyama, T. (1984). Staining of living bacteria with
rhodamine 123. FEMS Microbiology Letters 21, 153–7.
34. Mason, D. J. & Gant, V. A. (1995). The application of flow
cytometry to the estimation of bacterial antibiotic susceptibility
Journal of Antimicrobial Chemotherapy 36, 441–3.
31. Kaprelyants, A. S. & Kell, D. B. (1992). Rapid assessment of
bacterial viability and vitality by rhodamine 123 and flow cytometry.
Journal of Applied Bacteriology 72, 410–22.
32. Mason, D. J., Allman, R. & Lloyd, D. (1993). Uses of
membrane potential sensitive dyes with bacteria. In Flow Cyto metry in Microbiology (Lloyd, D., Ed.), pp. 67–81. Springer-Verlag,
London.
83
Received 11 July 1996; returned 25 October 1996; revised 9
December 1996; accepted 3 March 1997