Flow cytometry (FCM) is a technique for the rapid, opti-

cal analysis of individual cells. Measurements are made

by an array of detectors as the cells flow in a fluid stream
through a laser (or arc lamp) beam [Figure 1]. At the
sample interrogation point, light is scattered by the cells;
the extent of light scatter provides information on the
size and structure of the cell. In addition, fluorescence
may result from the absorption and re-emission of light
by chemicals that are either naturally present within the
cell (autofluorescence), or which have been added to the
sample prior to analysis.
FCM has many advantages over conventional cytometry.
Firstly, since acquisition rates of up to 10,000 cells.sec
-1
can be achieved (depending on the instrument used),
flow cytometric data sets often represent measurements
of in excess of 100,000 cells. In contrast, measurements
by microscopy often involve only a few hundred cells.
The increased sample throughput of FCM leads to the
acquisition of statistically significant results and the
detection of rare cell types. Secondly, since FCM uses
very sensitive electronic detectors to measure the intensi-
ty of scattered light or fluorescence at a given wavelength,
different intensities of light scatter/fluorescence can be
distinguished. By calibrating an instrument with samples
of known size or fluorescent intensity, it is possible to
obtain quantitative measurements. Thirdly, by using
dichroic filters to optically separate light of different
wavelength, flow cytometric measurements can be made
on several different characteristics of each cell. Typical
commercial flow cytometers allow 5-10 different param-
eters (e.g. size, protein content, DNA content, lipid con-
tent, antigenic properties, enzyme activity, etc.) to be col-
lected for each cell, allowing the operator to distinguish
between different cell types. Finally, since measurements
are made on single cells, heterogeneity within the popu-
lation can be detected and quantified in a way that can-
not be achieved by other means.
Whilst all commercial flow cytometers have the advan-
tages described above, some specialised instruments (cell
sorters) are able to physically separate cells on the basis of
user-defined characteristics. Depending on the instru-
ment, cells may be bulk-sorted or individual cells may be
sorted onto microscope slides or microtitre/agar plates.
Providing that appropriate cell staining and sample
preparation methods have been used to maintain viabil-
ity, sorted cells can be grown to give clonal colonies or
broth suspensions for con-
firmation of identity via
standard clinical microbi-
ology methods.
Over recent years a num-
ber of reviews of FCM
have been published [see
examples in reference 1].
The purpose of this review
is to highlight the value of
FCM for clinical samples,
with particular reference
to microorganisms.
Clinical applications of
microbial detection
by FCM
The detection of bacteria or
yeasts in body fluids is important for the diagnosis of a
number of different diseases. Urine may contain a variety
of particulates, including red and white blood cells, epithe-
lial cells, bacteria and inorganic chemical crystals. The
presence and concentrations of these particulates can be
used for the diagnosis of a range of diseases and disorders.
Flow cytometers designed specifically for urinalysis are
available commercially and these allow the simultaneous
determination of many different cell types [2]. These
devices have been shown to be more sensitive than manu-
al microscopic methods [3].
In comparison to the relatively straightforward detection
of bacteria in urine samples, blood is a much more chal-
lenging sample type to use. In clinical infections such as
bacteraemia, concentrations of the contaminants may be
of the order of 10 bacteria in 1 mL of blood, whilst the
number of red blood cells is >10
9
per mL. The high 'back-
ground' cellular load of blood makes the detection of bac-
teria by microscopic methods all but impossible.
Consequently, although bacteraemia is a potentially life-
threatening condition, diagnosis relies in many cases
upon the growth of bacteria in media inoculated with
samples of whole blood. However, methods are available
to selectively lyse the erythrocytes in a blood sample, leav-
ing a sufficiently low cell concentration to allow the rapid
sample throughput capabilities of the flow cytometer to
be utilised for the detection of bacteria. A number of
products are now available commercially to achieve this,
for example, CyLyse from Partec GmbH, Mnster,
Germany.
Mansour and colleagues [4] developed a model system in
which they used ethidium bromide labelling to non-
specifically detect Escherichia coli in blood at concentra-
tions of 10 - 100 cells.ml
-1
. The sensitivity was 100 to
1000-fold better than that achieved using microscopy
techniques, and took just 2 hours to perform, including
sample preparation. In clinical presentations where bac-
terial concentrations are less than 10 per mL, a short pre-
incubation step prior to flow cytometric analysis may be
envisaged to increase the bacterial load of the sample to a
level where it may be detected.
The detection of specific pathogenic microorganisms in
clinical samples has been much improved by the avail-
ability of monoclonal antibodies. These antibodies can
be fluorescently labelled (either directly or indirectly) to
enable them to be detected flow cytometrically. A variety
of fluoresecent labels are available, the most common is
fluorescein isothiocyanate (FITC). This has the advan-
tage of being well-excited by the 488 nm Argon ion laser
which is used as standard in most flow cytometers. Other
(spectrally-distinct) molecules such as allophycocyanin,
Texas Red and phycoerythrin allow multiple targets to be
detected simultaneously. The labelled-antibody approach
has proven to be useful for the detection of mycobacteri-
al species from clinical (sputum) specimens [5]. Yi and
colleagues showed that Mycobacteria could be detected
F low Cytometry
Flow cytometry for clinical
microbiology
as published in CLI February/March 2004
Flow cytometry (FCM) is a rapid technique for the analysis of individual cells. Light scattering and fluorescence properties of
cells are analysed as the cells pass through a laser beam and, in specialised instruments, cells with specific characteristics can
be isolated. This review article describes FCM and discusses recent advances that may be expected to increase its use in clin-
ical microbiology. New applications include susceptibility testing, where FCM allows death or damage to microorganisms to
be identified without the necessity to observe microbial growth, as well as monitoring the status and extent of infection in
HIV-positive patients.
by Dr. Hazel Davey
l
a
b

t
e
c
h
n
o
l
o
g
y
Dichroic
Filters
Flow
cell
Waste
Bandpass
Filters
Lasers & Lamps
Figure 1. Schematic drawing of a generalised flow cytometer. Modified with permission from
a drawing by Robert Murphy, Carnegie Melon University, Pittsburgh, PA, USA. (The Purdue
Cytometry CD-ROM Volume 4, J. Watson, Guest Ed., J. Paul Robinson, Publisher. Purdue
University Cytometry Laboratories, West Lafayette, IN, USA. 1997, ISBN 1-890473-03-0).
in as little as 3 hours; since Mycobacteria grow very slow-
ly in laboratory culture, a detection method that does not
rely on growth is very advantageous for clinical diagnos-
tic purposes. The method described used a rabbit poly-
clonal antibody against Mycobacterium species together
with a goat anti-rabbit IgG secondary antibody labelled
with R-phycoerythrin, and detected several different
Mycobacterium species. However, use of a species-specif-
ic antibody as the primary antibody would allow the
method to be used to detect M. tuberculosis specifically.
Susceptibility testing
In an era of worrying and increasing levels of antibiotic-
resistant pathogens, it is not surprising that understand-
ing the interactions between microorganisms and the
drugs designed to kill them has become another impor-
tant area for the clinical application of flow cytometric
methods. A variety of fluorescent
stains for assessing the viability of
microorganisms have been identi-
fied [Table 1, see also reference 6]
and these are particularly useful
for determining the efficacy of
antimicrobial compounds.
Microorganisms exposed to
antibiotic or antifungal com-
pounds (either in vivo or in vitro)
are compared to control (untreat-
ed) samples and appropriate
stains are used to identify changes
in nucleic acids, proteins, mem-
branes, etc.
Antibiotics disrupt cellular activi-
ties and the particular mode of
action can be determined flow
cytometrically. For example,
antibiotic-induced damage to cell membranes can be
detected by the entry of fluorescent compounds (such as
propidium iodide) which are normally excluded by the
intact cell membrane. Alternatively, to deter-
mine the response of cells to an antibiotic,
which affects nucleic acid synthesis, one could
use a stain such as DAPI, which binds to DNA,
or pyronin Y, which binds to RNA.
In addition, FCM permits subpopulations with
varying resistance to be identified and accurate
assessment of the dose-response curve can also
be performed as part of the assay [see examples
in reference 7]. Flow cytometric susceptibility
testing thus allows death or damage of microor-
ganisms to be identified without the necessity to
observe microbial growth (or lack thereof).
Flow cytometric susceptibility testing can be
performed in a few hours [Figure 2] and conse-
quently this method has the potential to con-
tribute to the decision of which drug or drug
combination would be most appropriate for a
particular patient.
HIV
FCM has been used to great effect for monitor-
ing the status and extent of HIV infection. Whilst
viral antigens can be detected by FCM [8], monitoring of
HIV infection usually relies on regular quantitation of
lymphocyte populations. The absolute numbers of CD4+
lymphocytes and their percentage values within the total
lymphocyte populations are good indicators of the dis-
ease and its progression. Fluorescently-labelled antibod-
ies can be used to selectively label different types of lym-
phocytes and thus FCM has an important role to play not
only in disease surveillance, but also in determining the
efficacy of treatment. Ideally analysis of blood samples
should be performed within hours of collection.
Unfortunately, the majority of HIV-infected individuals
are not within easy reach of the specialised laboratories
capable of performing these tests. A mobile flow cytom-
etry laboratory has recently been developed to address
this issue (Partec GmbH, Mnster, Germany). The
CyFlow flow cytometer is installed in an off-road 4-wheel
drive car and is powered using 12 V DC car batteries
charged by solar panels [Figure 3]. The system has advan-
tages over many flow cytometers in that lymphocyte pop-
ulations can be simultaneously identified and quantified
without the addition of reference controls [9]. Detection
of the different lymphocyte populations is achieved using
monoclonal antibodies targeted against the appropriate
CD markers. The cells in a fixed volume (200 mL) of
sample are counted; counting is switched on and off
using an electrode to sense the depth of fluid in the sam-
ple tube. The combined detection and counting not only
simplifies the procedure, thus reducing the potential for
error, but also minimises costs.
Future prospects
A recent development that may be expected to promote
the use of FCM for the analysis of clinical samples is the
Amnis ImageStream System (www.amnis.com), which
permits images of individual cells to be captured along
with their multiparametric flow cytometric data. Thus
dots on a flow cytometric data plot can be directly linked
to an image of the cell. This has particular use when
"abnormal" signals are detected by FCM - the operator
can relate these signals back to up to six separate images
of the cell to check for the presence of cell doublets, con-
taminating cell types or to verify the result of screening
tests.
Over the last few years, kits designed specifically for the
flow cytometric analysis of microorganisms have become
available (see e.g. www.bdbiosciences.ca /downloads/hot-
lines/Cell_Viability_HL_Fall2003.pdf and
www.probes.com/ handbook/sections/1503.html). The
growing popularity of such kits reflects, at least in part,
their ease of use. Whilst this is to be welcomed, there is
some danger that the kits may be adopted without analysis
of proper control standards. Despite the names of these
kits, distinguishing live and dead bacteria and yeasts is not
always straightforward and care in interpretation of the
results is still of great importance.
In conclusion, FCM offers many advantages for clinical
microbiology. Recent developments are likely to open up
further possibilities of new applications, as well as increas-
ing the use of existing flow cytometric techniques.
as published in CLI February/March 2004
F low Cytometry
Stain
BacLight Kit: Molecular
Probes www.probes.com
bis-(1,3-dibutylbarbituric
acid) trimethine oxonol
(DiBAC4(3))
Calcofluor White
5-cyano-2,3-ditolyltetra-
zolium chloride (CTC)
Fluorescein diacetate/
Carboxy-fluorescein
diacetate
Rhodamine 123
TO-PRO-3 / Propidium
iodide
Mode of Action
Propidium iodide excluded
by intact membranes. All
cells take up SYTO9
Uptake by dead cells
Uptake by dead cells
Respiratory activity
Enzymic activity
Uptake by live cells
Excluded by intact cell
membrane
Results
Live cells are green, dead
cells are red.
Dead cells appear
green/yellow.
Dead cells appear blue.
Live cells appear red.
Live cells appear green.
Live cells appear green.
Dead cells appear red.
Table 1. Some fluorescent dyes used for determination of viability by FCM.
Untreated 40 min.
1 hour 3 hours
Red Fluorescence
G
r
e
e
n

F
l
u
o
r
e
s
c
e
n
c
e
G
r
e
e
n

F
l
u
o
r
e
s
c
e
n
c
e
G
r
e
e
n

F
l
u
o
r
e
s
c
e
n
c
e
G
r
e
e
n

F
l
u
o
r
e
s
c
e
n
c
e
Red Fluorescence
Red Fluorescence Red Fluorescence
Figure 2. Antimicrobial susceptibility testing using flow
cytometry. Two colour fluorescence histograms of
Enterococcus faecium treated with vancomycin and
stained with the FAST-2 kit (BioRad). With increasing
exposure time, an increase in the number of dead and
dying cells (events present in quadrants 2, 3, and 4) was
observed. Data collected by Kuo-Ping Chiu and colleagues
at BioRad, printed with permission (The Purdue
Cytometry CD-ROM Volume 4, J. Watson, Guest Ed., J.
Paul Robinson, Publisher. Purdue University Cytometry
Laboratories, West Lafayette, IN, USA. 1997, ISBN 1-
890473-03-0).
Figure 3. The CyFlow flow cytometer, image kindly provided by
Partec, GmbH.
References
1. Davey HM, Kell DB. Flow cytometry and cell sorting of
heterogeneous microbial populations-the importance of
single-cell analyses. Microbiological reviews
1996;60(4):641-696.
2. Delanghe JR, Kouri TT, Huber AR, Hannemann-Pohl
K, Guder WG, Lun A, Sinha P, Stamminger G, Beier L. The
role of automated urine particle flow cytometry in clini-
cal practice. Clinica Chimica Acta 2000;301(1-2):1-18.
3. Hannemann-Pohl K, Kampf SC. Automation of urine
sediment examination: A comparison of the sysmex UF-
100 automated flow cytometer with routine manual diag-
nosis (microscopy, test strips, and bacterial culture).
Clinical Chemistry and Laboratory Medicine
1999;37(7):753-764.
4. Mansour JD, Robson JA, Arndt CW, Schulte TE.
Detection of Escherichia coli in blood using flow cytome-
try. Cytometry 1985;6:186-190.
5. Yi WC, Hsiao S, Liu JH, et al. Use of fluorescein labelled
antibody and fluorescence activated cell sorter for rapid
identification of Mycobacterium species. Biochem
Biophys Res Commun 1998;250(2):403-8.
6. Davey HM, Kaprelyants AS, Weichart DH, Kell DB.
Estimation of microbial viability using flow cytometry.
Current Protocols in Cytometry. New York: Wiley; 1999. p
11.3.1-11.3.20.
7. Pore RS. Ketoconazole susceptibility of yeasts by the
FCST method. Current Microbiol. 1991;23:45-50.
8. McSharry JJ. Uses of flow cytometry in virology.
Clinical microbiology reviews 1994;7(4):576.
9. Greve B, Cassens U, Westerberg C, Jun WG, Sibrowski
W, Reichelt D, Gohde W. A new no-lyse, no-wash flow-
cytometric method for the determination of CD4 T cells
in blood samples. Transfusion Medicine and
Hemotherapy 2003;30(1):8-13.
The author
Hazel M. Davey, Ph.D.,
Postdoctoral Research Assistant,
Institute of Biological Sciences, University of Wales,
Aberystwyth, Ceredigion, SY23 3DD,
Wales, U.K.
Tel.: +44 1970 621829
Fax: +44 1970 622307
Email: hlr@aber.ac.uk
Website: http://qbab.aber.ac.uk/home.html
as published in CLI February/March 2004
F low Cytometry