Development of a simple model for studying the effects of antifungal agents on multicellular communities of Aspergillus fumigatus

Abstract

Aspergillus fumigatus is an increasingly prevalent opportunistic fungal pathogen of various immunocompromised individuals. It has the ability to form filaments within the lungs, producing dense intertwined mycelial balls, which are difficult to treat. The aim of this study was to develop a suitable model of A. fumigatus to examine the effects of antifungal challenge on these intertwined filamentous communities. A. fumigatus NCPF 7367 growth conditions were optimized on both Thermanox coverslips and on flat-bottomed microtitre plates to establish optimal conidial seeding densities. Isolates were treated with itraconazole, voriconazole, amphotericin B and caspofungin and their overall killing efficiency was measured using an XTT formazan metabolic dye assay. This was compared with the CLSI (formerly NCCLS) methodology of broth microdilution of moulds (standard M38-A). It was shown that 1x10(5) conidia ml(-1) in RPMI 1640 was the optimum concentration of spores for biofilm formation. Filamentous growth characteristics were not observed until 10 h incubation, followed by an exponential increase in the biofilm biomass (hyphae and extracellular material) and cellular activity (metabolism). When susceptibility testing of biofilms was compared with that of planktonic cells by CLSI broth microdilution testing, all antifungal drugs were at least 1000 times less effective at reducing the overall metabolic activity of 90 % of the cells. Overall, this study showed that A. fumigatus has the ability to form coherent multicellular biofilm structures that are resistant to the effects of antifungal drugs.

Full text

Development of a simple model for studying the
effects of antifungal agents on multicellular
communities of Aspergillus fumigatus
Eilidh Mowat,
1
John Butcher,
1
Sue Lang,
1
Craig Williams
2
and Gordon Ramage
3
Correspondence
Gordon Ramage
g.ramage@dental.gla.ac.uk
1
Department of Biological and Biomedical Sciences, Glasgow Caledonian University, Glasgow,
UK
2
Microbiology Department, Yorkhill Hospital, Glasgow, UK
3
Section of Infection and Immunity, Glasgow University Dental School and Hospital, Glasgow, UK
Received 21 February 2007
Accepted 14 May 2007
Aspergillus fumigatus is an increasingly prevalent opportunistic fungal pathogen of various
immunocompromised individuals. It has the ability to form filaments within the lungs, producing
dense intertwined mycelial balls, which are difficult to treat. The aim of this study was to develop a
suitable model of A. fumigatus to examine the effects of antifungal challenge on these intertwined
filamentous communities. A. fumigatus NCPF 7367 growth conditions were optimized on both
Thermanox coverslips and on flat-bottomed microtitre plates to establish optimal conidial seeding
densities. Isolates were treated with itraconazole, voriconazole, amphotericin B and caspofungin
and their overall killing efficiency was measured using an XTT formazan metabolic dye assay.
This was compared with the CLSI (formerly NCCLS) methodology of broth microdilution of
moulds (standard M38-A). It was shown that 1¾10
5
conidia ml
”1
in RPMI 1640 was the optimum
concentration of spores for biofilm formation. Filamentous growth characteristics were not
observed until 10 h incubation, followed by an exponential increase in the biofilm biomass (hyphae
and extracellular material) and cellular activity (metabolism). When susceptibility testing of biofilms
was compared with that of planktonic cells by CLSI broth microdilution testing, all antifungal
drugs were at least 1000 times less effective at reducing the overall metabolic activity of 90 % of
the cells. Overall, this study showed that A. fumigatus has the ability to form coherent multicellular
biofilm structures that are resistant to the effects of antifungal drugs.
INTRODUCTION
Aspergillus species are opportunistic filament-forming
moulds. The genus comprises over 180 species, with
Aspergillus fumigatus causing the majority of human
aspergillus infections (Hope et al., 2005). A. fumigatus is
a ubiquitous saprophytic fungus w ith a worldwide
distribution due to the production of small spores called
conidia, which have a mean size of 2–3.5
mm, resulting in
the conidia being dispersed into the air and remaining in
the atmosphere for time periods that depend on the
conditions of the external environment, such as tempera-
ture, humidity and seasonal variations (Rivera et al., 2006).
In recent times, the frequency of disseminated fungal
infections has increased dramatically. Overall, A. fumigat us
is now the second most common cause of fungal infection
found in hospitalized patients, after Candida albicans (Ellis
et al., 2000). Aspergillus conidia are usually eliminated
efficiently by the innate and acquired immune systems.
However, in immunocompromised pati ents, such as
transplant, leukaemia and human immunodeficiency virus
(HIV)-positive patients, A. fumigatus can cause a range of
systemic diseases with mortality rates ranging from 30 to
90 % (Brakhage, 2005; Den ning et al., 1998; Herbrecht
et al., 2002). Pulmonary infection may also occur in other
patients such as those with cystic fibrosis (CF). In thes e
patients, infection with A. fumigatus may cause allergic
bronchopulmonary aspergillosis (ABPA), a mycetoma
(fungus ball) or invasive aspergillosis (IA) (de Almeida
et al., 2006; Shibuya et al., 2004).
The initial establishment of chronic A. fumigatus infection
involves the germination of conidia and subsequent hyphal
invasion of the lung tissue (Filler & Sheppard, 2006).
Histological and microscopic examination of bronchopul-
monary lavage samples has revealed the presence of
Abbreviations: ABPA allergic bronchopulmonary aspergillosis; AmpB,
amphotericin B; CF, cystic fibrosis; CLSM, confocal laser scanning
microscopy; IA, invasive aspergillosis; PMIC, planktonic cell MIC; SMIC,
sessile cell MIC; XTT, 2,3-bis(2-methoxy-4-nitro-5-sulfo-phenyl)-2H-
tetrazolium-5-carboxanilide.
Journal of Medical Microbiology (2007), 56, 1205–1212 DOI 10.1099/jmm.0.47247-0
47247
G
2007 SGM Printed in Great Britain 1205

numerous A. fumigatus hyphae in the form of a complex
multicellular structure (mycetoma), which is similar to the
biofilms formed by Candida species (Ramage et al., 2001c,
2005). In contrast to the biofilms formed by Candida
species, very limited information is currently available on
the development and behaviour of A. fumigatus adherent
multicellular communities and their response to antifungal
treatment. To date, there are only two reports suggesting
that Aspergillus species are able to grow and form biofilms
(Beauvais et al., 2007; Villena & Gutierrez-Correa, 2006).
The purpose of this study was to investigate the growth
characteristics of filamentous A. fumigatus multicellular
communities through the development of an in vitro model
that could be utilized to screen the growth characteristics of
clinical isolates or mutants, and to examine the antifungal
susceptibility profiles of complex biofilm-like structures
(Ramage et al., 2002a, c).
METHODS
Organisms. A. fumigatus NCPF 7367 and four clinical isolates
(YHCF1, YHCF2, YHCF3 and YHCF4) obtained from the Royal
Hospital for Sick Children (Yorkhill Division, NHS Glasgow, UK)
were used throughout this study. All isolates were stored on
Sabouraud dextrose agar slopes (Oxoid) at 4 uC.
Growth conditions and standardization of conidial inoculum. A.
fumigatus was grown on Sabouraud dextrose agar at 37 uC for 72 h.
Conidia were harvested by flooding the surface of the agar plates with
5 ml PBS (Oxoid) containing 0.025 % (v/v) Tween 20 and rocking
gently. The conidial suspension was recovered and dispensed into a
5 ml sterile glass bottle. The conidia were counted using a Neubauer
haemocytometer and adjusted to the required concentration in RPMI
1640 (Sigma) buffered to pH 7.0 with 0.165 M MOPS. All procedures
were carried out in a HERASafe laminate flow cabinet (model K515;
Kendro).
Biofilm formation. A. fumigatus biofilms were formed on commer-
cially available, pre-sterilized, polystyrene, flat-bottomed, 96-well
microtitre plates (Corning). Biofilms were formed by adding 200
mlof
a standardized cell suspension in MOPS-buffered RPMI 1640 to each
well for selected time periods (4, 8, 12, 24 and 48 h), and incubating
statically at 37 uC. A minimum of 12 replicates was performed for
each experimental parameter, plus suitable controls. At each selected
time point, the medium was aspirated and the biofilms were washed
thoroughly three times with sterile PBS by repeated pipetting to
remove non-adherent cells.
Confocal laser scanning microscopy (CLSM). A. fumigatus
biofilms were formed as described above on the surface of 13-mm-
diameter sterile Thermanox plastic cell culture coverslips (Nunc) in
24-well tissue culture plates (Nunc). After incubation at 37 u C for
various time periods (0, 2, 4, 8, 10, 12, 16, 18 and 24 h), the coverslips
were washed in sterile PBS and stained using the LIVE/DEAD
fluorescent stain (Molecular Probes), according to the manufacturer’s
instructions. The FUN1 component of the kit was used, which is a
bright green fluorescent intracellular stain. This was applied to the
washed biofilms for 20 min in the dark. The biofilms were then
washed in PBS and mounted on a slide. The fluorescent filamentous
biomass was examined using a Zeiss Axiovert LSM510 confocal
microscope attached to an LSM510 laser scanning system with a
488 argon ion laser at 6200 magnification. Sections of the xy plane
were taken at 1
mm intervals along the z-axis to determine the depth
of the biofilms and overall physical ultrastructure. Three-dimensional
images were obtained using computer software.
Biofilm quantification. Biofilm biomass was assessed using a
modified version of a protocol first developed by Christensen et al.
(1985) and subsequently modified by O’Toole & Kolter (1998). At
each time interval, the spent culture medium was removed from each
well and the adherent cells were washed three times with PBS. These
were air-dried and 100
ml of 0.5 % (w/v) crystal violet solution was
added for 5 min. The solution was then removed by carefully rinsing
the biofilms under running water until excess stain was removed. The
biofilms were destained by adding 100
ml 95 % ethanol to each well.
The ethanol was gently pipetted to completely solubilize the crystal
violet for 1 min, the ethanol was transferred to a clean 96-well
microtitre plate and the A
570
was read (FLUO Star Optima
fluorescence microplate reader; BMG Labtech). The absorbance
values are proportional to the quantity of biofilm biomass, which
comprises hyphae and extracellular polymeric material (the greater
the quantity of biological material, the higher the level of staining and
absorbance).
2,3-bis(2-Methoxy-4-nitro-5-sulfo-phenyl)-2H-tetrazolium-5-
carboxanilide (XTT) reduction assay.
A semi-quantitative measure
of each biofilm was calculated using an XTT reduction assay, adapted
from previous studies (Hawser et al., 1998). This is a metabolic
reduction assay that measures the activity of cells and can be used to
compare untreated cells with cells treated with antimicrobial agents,
and has been used previously in studies to evaluate the antifungal
sensitivities of filamentous fungi (Antachopoulos et al., 2006;
Meletiadis et al. , 2001b, c). Briefly, XTT (Sigma) was prepared in a
saturated solution at 0.5 g l
21
in PBS. The solution was filter-
sterilized through a 22
mm pore size filter, aliquoted and stored at
280 uC. Prior to each assay, an aliquot of stock XTT was thawed and
menadione (10 mM prepared in acetone; Sigma) was added to a final
concentration of 10
mM. A 100 ml aliquot of XTT/menadione solution
was added to each well and to appropriate control wells to measure
background XTT reduction levels. The plates were incubated in the
dark for 3 h at 37 uC and the colour change was measured using a
490 nm filter in a microplate reader (FLUO Star Optima; BMG
Labtech). The colorimetric change in the XTT reduction assay directly
correlates with the metabolic activity of the biofilm.
End-point susceptibility testing
M38-A broth microdilution testing.
Planktonic cell MICs (PMICs)
were evaluated using the CLSI (formerly NCCLS) M38-A standard
methodology (NCCLS, 2002). The antifungal agents itraconazole
(Sigma), voriconazole (Pfizer), amphotericin B (AmpB; Bristol Myers
Squibb) and caspofungin (Merck) were prepared as stock solutions in
DMSO (Sigma) and diluted in MOPS-buffered RPMI 1640 to
working concentrations. Microtitre plates containing 100
ml of each
antifungal were serially diluted twofold to produce a concentration
range of 0.03–16 mg l
21
. A range of conidial suspensions was then
prepared in MOPS-buffered RPMI 1640 to a final concentration of
0.4–5610
4
conidia ml
21
and 100 ml was added to each well. The
plates were incubated for 48 h at 35 uC. The PMIC for each antifungal
was defined as the lowest concentration that produced complete
visible inhibition of growth. Testing of these isolates was performed in
quadruplicate.
Modified M38-A broth microdilution testing. For antifungal
susceptibility testing of biofilms
[
sessile cell MICs (SMICs)
]
, conidia
were prepared as described above and standardized to a density of
1610
6
conidia ml
21
in MOPS-buffered RPMI 1640. Biofilms were
formed by pipetting standardized conidial suspensions into selected
wells of a microtitre plate and incubating for 24 h at 35 uC, as
E. Mowat and others
1206 Journal of Medical Microbiology 56

described above. After growth, the medium was aspirated and non-
adherent cells were removed by thorough washing of the cells (three
times) using sterile PBS and gentle pipetting. Residual PBS from each
well was removed by blotting with paper towels. All antifungals
(itraconazole, voriconazole, AmpB and caspofungin) were prepared
as described above to provide a working concentration of 512 mg l
21
in MOPS-buffered RPMI 1640. These were serially diluted twofold
(1–256 mg l
21
) directly into adjacent wells and the challenged cells
were incubated statically for a further 48 h at 35 uC. A number of
antifungal-free wells and biofilm-free wells were also included to
serve as positive and negative controls, respectively. SMICs were
determined as 50 and 90 % reduction in metabolism compared with
the untreated control using the XTT reduction assay described above.
Testing of these isolates was performed in quadruplicate.
Statistical analysis. The absorbance values of individual biofilms
were compared by one-way analysis of variance and using the
Bartlett’s test for homogeneity of variances and the Bonferroni’s
multiple comparison post-test. A value of P ,0.05 was considered to
be significant. Analyses were performed using
SPSS 13.0 for Windows.
RESULTS AND DISCUSSION
A. fumigatus is now a leading fungal pathogen and one of
the most significant opportunistic fungi in bone-marrow-
transplant patients (Singh, 2005). A. fumigatus is also
found in a range of other patient groups, including CF
patients, HIV-positive patients and other immunocom-
promised individuals (Cimon et al., 2001). The presence of
mycetomas in the upper airways, the pulmonary epithelial
cells of the alveoli or in the maxillary sinuses provides
compelling evidence that A. fumigatus biofilms are more
prolific and problematic than once thought (Filler &
Sheppard, 2006; Mensi et al., 2004). These infections are
typified by intricate networks of hyphae that develop from
inhaled conidia (Shibuya et al., 2004). Morphologi cally,
these structures resemble other fungal biofilms, such as
those of C. albicans (Ramage et al., 2001c), which are
clinically important due to their role in pathogenesis and
resistance to antifungal agents (Ramage et al., 2002c, 2005).
Therefore, we hypothesized that similar pathogenic traits
might be exhibited by this fungus, meriting further
investigation.
Standardizing an in vitro model
To date, there have been no published accounts establish-
ing that filamentous A. fumigatus grows as a biofilm.
Histopathological evidence has indicated the presence of
mycelial plugs and proliferation of hyphae with acute-angle
dichotomous branching (Shibuya et al., 2004), which are
complicit with the definition of a biofilm. There are many
forms of aspergillosis, ranging from ABPA to IA, and
despite different clinical presentations, morphological
characteristics remain relatively similar and are typified
by intricate mycelial networks. Formation of A. fumigatus
mycelial aggregates that exhibit classic biofilm character-
istics is reported here, i.e. an adherent microbial popula-
tion, adherent to each other and/or surfaces or interfaces
(Costerton et al., 1995). To our knowledge, this study is the
first to examine submerged cultures of A. fumigatus in vitro
in this growth modality.
A key factor that we noticed early in our studies was the
critical importance of conidial seeding density. Clearly, the
structural morphology and integrity of these multicellular
structures was dependent on the concentration of conidia
(ml medium)
21
(Figs 1 and 2), a phenomenon previously
identified with C. albicans biofilm development (Ramage
et al., 2001c). To produce adherent aggregated multi-
cellular populations (biofilms) with a similar morphology
to those seen during in vivo aspergillosis lung infecti on, we
developed a model, initially examining conidial seeding
density of A. fumigatus NCPF 7367 and four clin ical
isolates. First, we examined a serial dilution of conidial
densities, ranging from 10 to 1610
6
conidia ml
21
, and
examined the resultant multicellular structures after a 24 h
inoculation using semi-quantitative metabolic and biomass
assays. Fig. 1 illustrates that both the metabolic activity and
the biomass of the biofilms exhibited a positive correlation
with the conidial seeding density, although at the highest
concentrations of conidia the biomass appeared to decline.
We subsequently used CLSM, a non-invasive technique
that enabled three-dimensi onal structural imaging and
depth measurements of intact multicellular structures on
1.6
XTT A
490
Crystal violet A
570
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1×10
6
1×10
5
1×10
4
1×10
3
Conidia inoculum ml
_
1
(a)
(b)
1×10
2
1×10
1
1×10
6
1×10
5
1×10
4
1×10
3
Conidia inoculum ml
_
1
1×10
2
1×10
1
Fig. 1. Effect of conidial concentration on A. fumigatus NCPF
7367 biofilm development after 24 h. Metabolic activity of the
biofilm (a) and the biomass (b) increased with the concentration of
conidia seeded. Note how there is a limit to the development of the
biofilm, as indicated from the biomass generated from an inoculum
of 1¾10
6
conidia ml
”1
.
Aspergillus fumigatus biofilms
http://jmm.sgmjournals.org 1207

Thermanox coverslips. The mature biofilms formed using a
series of standardized conidial suspensions (1610
4
,1610
5
and 1610
6
conidia ml
21
) were evaluated by microscopy to
investigate the optimal conidial inoculum concentration.
Fig. 2(a) illustrates complex multicellular aggregates
formed from three different conidial seeding densities after
growth for 24 h, which rang ed in depth from 117 to
360
mm. All of the filamentous multicellular structures
exhibited acute-angle dichotomous bran ching to a varying
extent. However, the conidial seeding density played an
important role in the overall structural integrity of the
biofilm structure. Biofilm stability was assessed by shear
mechanical force by serially pipetting the biofilms with PBS
during the w ashing procedure (results not shown). Either
increasing or decreasing the conidial seeding density 10-
fold led to significant differences in the depth of the
resultant bio films (P .0.05). At the highest conidial
concentration (1610
6
conidia ml
21
), the biofilms were
relatively thin (117
mm) and development of filamentous
growth was severely restricted, resulting in less overall
biomass (Fig. 2). Conversely, biofilms formed by fewer
seeded conidia (1610
4
conidia ml
21
) were observed to be
thicker (360
mm) and have longer mycelial frameworks.
These were easily disrupted and removed by mechanical
forces and were therefore not reproducible. The conidial
concentration of 1610
5
conidia ml
21
produced robust
filamentous structures that were resistant to mechanical
disruption (Fig. 2). Therefore, this concentration of
conidia was selected, as the resultant biofilms exhibited
reproducible characteristics that were amenable to the
high-throughput testing required for screening of clin ical
isolates and defined mutants, or for testing the suscept-
ibility of antifungal agents.
Kinetics of biofilm development
The growth and development of A. fumigatus in the lung,
from initial inhalation of conidia to filamentous forms
associated with mycetomas, is a key factor in pathogenesis.
Clearance of inhaled conidia via innate immune mechan-
isms in immunocompetent individuals is the key defence
against aspergillosis (Latge, 2001). The progressi on of
disease in susceptible patient groups is dependent on the
ability of the conidia to germinate and form mycelial
masses (mycetomas), which then penetrate the pulmonary
epithelium prior to angioinvasion and systemic spread
(Filler & Sheppard, 2006; Kamai et al., 2006). In our
studies, we examined the kinetics of multicellular devel-
opment. Initial adherence, conidial germination, and
filamentous growth and differentiation of A. fumigatus
1×10
4
conidia ml
_
1
1×10
5
conidia ml
_
1
1×10
6
conidia ml
_
1
(a)
(b)
450
*
*
*
Biofilm depth (mm)
400
350
300
250
200
150
100
50
0
10
4
10
5
Conidia inoculum ml
_
1
10
6
Fig. 2. CLSM of A. fumigatus NCPF 7367
biofilms generated using different conidial
concentrations. (a, b) CLSM of biofilms formed
by various conidial seeding concentrations on
Thermanox coverslips. Differences in structural
integrity associated with different conidial
density are shown in (a). Higher conidial
inoculum concentrations were associated with
poorer filamentation and biofilm structural
integrity. An inverse relationship was found
between conidial density and biofilm depth (b).
Significant differences were found between
the biofilm depths (*, P ,0.05).
Fig. 2 CLSM of A. fumigatus NCPF 7367 biofilms generated
using different conidial concentrations. (a, b) CLSM of biofilms
formed by various conidial seeding concentrations on Thermanox
coverslips. Differences in structural integrity associated with
different conidial density are shown in (a). Higher conidial inoculum
concentrations were associated with poorer filamentation and
biofilm structural integrity. An inverse relationship was found
between conidial density and biofilm depth (b). Significant
differences were found between the biofilm depths (*, P ,0.05).
E. Mowat and others
1208 Journal of Medical Microbiology 56

NCPF 7367 and four clinical isolates were analysed on the
surface of polystyrene wells over 48 h (1610
5
con-
idia ml
21
), as assessed by metabolic activity (XTT) and
biomass (crystal violet) measurements (Fig. 3). A correla-
tion between metabolism, biomass and hyphal develop-
ment was demonstrated. The metabolic activity and the
biomass of the biofilms were shown to increase over time.
Microscopic analysis revealed that A. fumigatus did not
form hyphae until approx imately 8 h, in agreement with
the results of Meletiadis et al. (2001a), who demonstrated
germination at approximately 6 h using a si milar micro-
titre plate method while examining filamentous growth
characteristics of A. fumigatus. When the kinetics of
biofilm formation was measured, their results were
comparable to those in this study, with a lag phase of
approximately 10 h. Hyphae began to intertwine forming a
monolayer (10–16 h), followed by increased structural
complexity over the subsequent 4–8 h (Fig. 4a). This was
shown more clearly from the depth measurements taken
over a 24 h period. Fig. 4(b) showed an exponential rise in
the depth of the biofilm between 10 and 18 h, which then
reached a plateau as development ceased and a steady state
occurred. These biofilm development characteristics were
observed for all isolates tested in this study. The
development of the biofilm was slower than that of C.
albicans biofilms; nevertheless, the overall characteristics of
development were similar (Chandra et al., 2001; Ramage
et al., 2001c) . The slower initial development of mycelia
may be related to the overall density of the cells, indicating
a potential role for quorum sensing in multicellular A.
fumigatus populations. This phenomenon has been
demonstrated in C. albicans with the secreted molecule
farnesol, which prevents hyphal development and sub-
sequent biofilm formation (Ramage et al., 2002b).
Antifungal susceptibility testing
Currently, antifungal therapy remains the main w ay of
controlling the progression of aspergillosis. It is clear from
our results that the activity of antifungal agents in vitro is
considerably different when biofilms are compared with
planktonic cells, and that a number of currently prescribed
antifungal agents may be ineffective in the treatment of
established biofilm-associated infections. The method-
ologies used to examine PMICs and SMICs are quite
disparate, i.e. there was a 3 log difference in the number of
conidia used in the initial inoculum between these assays.
This factor may account, in part, for the disparity between
the PMICs and SMICs reported here. For example,
itraconazole and caspofungin were ineffective against
multicellular structures, exhibiting over 1000-fold more
resistance than their planktonic counterparts (Table 1).
The current gold standard for choosing the most appro-
priate treatment for aspergillosis infections is the CLSI
standard M38-A broth microdilution in vitro antifungal
susceptibility testing assay (NCCLS, 2002). This assay
measures inhibition of growth and provides an accurate
method for the prediction of fungistatic antifungal drugs.
However, in this study we were unable to quantify the
effects of the antifungal drugs using this cri terion on
multicellular structures, which are no longer actively
dividing yet are representative of an established infection.
ABPA infections with a characteristic mycetoma are
typified by greater cellular burdens than are used in
planktonic assays and exh ibit biofilm characteristics. We
used a metabolic dye that exhibited a direct correlation
with cellular viability. Viable cell counting was not used
due to the inability to correlate colony counts with
individual cells. The assay that we described here provides
an accurate means of predicting in vivo activity due to the
reduced metabolic activity of the cells tested. XTT has been
used in other mycological studies to monitor antifungal
exposure (Antachopoulos et al., 2006; Hawser et al., 2001;
Ramage et al., 2001a). For example, Hawser et al. (2001)
previously described an XTT-based methodology for
measuring the minimum effective concentrations of anti-
fungal agents against A. fumigatus. In addition,
Antachopoulos et al. (2006) recently demonstrated how
this simple technology could be used to assess antifungal
activity against zygomycetes, and demonstrated a reprodu-
cible correlation between XTT-deduced MICs and the CLSI
methodology. Ramage and colleagues also showed its utility
against both fungal (C. albicans) and bacterial (Pseudomonas
aeruginosa
) cells, planktonic and sessile cells, and with
1.6
XTT A
490
1.4
1.2
1.0
0.8
0.6
0.4
0.2
10 20 30
Time (h)
NCPF 7367
YHCF1
YHCF2
YHCF3
YHCF4
40 50 60
Fig. 3. Growth kinetics of A. fumigatus NCPF
7367 and four CF isolates over a 48 h period
(seeding density 1¾10
5
conidia ml
”1
).
Aspergillus fumigatus biofilms
http://jmm.sgmjournals.org 1209

various antimicrobial agents (Ramage et al., 2001a; Tunney
et al., 2004). There may be limitations to this assay, such as
viable but non-culturable organisms. However, it is unlikely
that a high proportion of the biofilm organisms are
composed of these viable but non-culturable cells. Overall,
the assay enables the user to assess rapidly, either by naked
eye or by spectroscopic analysis, the concentration of
antifungal agent required to kill or inhibit A. fumigatus.
Table 1. Biofilm susceptibility testing
Concentrations are given as mg l
21
.
Isolate Itraconazole Voriconazole AmpB Caspofungin
PMIC SMIC
50
SMIC
90
PMIC SMIC
50
SMIC
90
PMIC SMIC
50
SMIC
90
PMIC SMIC
50
SMIC
90
NCPF 7367 0.5 .256 .256 0.25 16 .256 1.0 0.125 32 0.25 128 .256
YHCF1 0.5 .256 .256 0.25 16 .256 1.0 0.125 16 0.25 128 .256
YHCF2 0.5 .256 .256 1.0 128 .256 0.25 ,0.125 8 0.25 64 .256
YHCF3 0.5 .256 .256 0.5 32 .256 1.0 ,0.125 32 0.25 64 .256
YHCF4 0.25 .256 .256 0.25 16 .256 0.5 1 16 0.25 64 .256
4 h 8 h 10 h
16 h 18 h 24 h
250
(b)
(a)
200
150
100
50
51015
Time (h)
Biofilm depth (mm)
20 25 30
Fig. 4. CLSM of A. fumigatus NCPF 7367
biofilm growth kinetics with an inoculum of
1¾10
5
conidia ml
”1
. (a) Increasing levels of
filamentation and complexity were proportional
to time. (b) Biofilm depth was minimal for the
first 8–10 h, and subsequently became thicker
and more complex from 10–16 h, finally
maximizing at 24 h.
Fig. 4. CLSM of A. fumigatus NCPF 7367 biofilm growth kinetics
with an inoculum of 1¾10
5
conidia ml
”1
. (a) Increasing levels of
filamentation and complexity were proportional to time. (b) Biofilm
depth was minimal for the first 8–10 h, and subsequently became
thicker and more complex from 10–16 h, finally maximizing at 24 h.
E. Mowat and others
1210 Journal of Medical Microbiology 56

In this study, five strains of A. fumigatus were examined for
their susceptibility to a range of antifungal agents when
grown as a complex mycelial structure, using a method-
ology previously employed for C. albicans biofilms
(Ramage et al., 2001a). These agents included two azoles
(itraconazole and voriconazole), a polyene (AmpB) and an
echinocandin (caspofungin), which were serially diluted
and used to challenge planktonic and sessile cells. Table 1
illustrates the results from both conventional planktonic
CLSI M38-A susceptibility testi ng and the modified sessile
susceptibility testing. The results indicated that caspo-
fungin had a PMIC of 0.25 mg l
21
, itraconazole 0.25–
0.5 mg l
21
, voriconazole 0.25–1 mg l
21
and AmpB
0.25–1 mg l
21
. Testing of the sessile cells showed that
AmpB was the most effective antifungal agent, with SMIC
50
values ranging from ,0.125 to 1 mg l
21
and SMIC
90
values from 8 to 32 mg l
21
. The other azole, voriconazole,
exhibited increas ed efficacy (SMIC
50
of 16–128 mg l
21
).
Caspofungin showed poor overall activity against adherent
multicellular A. fumigatus, with SMIC
50
and SMIC
90
values
of 64–128 and .256 mg l
21
, respectively. Itraconazole was
ineffective against all adherent multicellular populations
(SMIC
50
and SMIC
90
values of .256 mg). Overall, AmpB
was the most effective against sessile cells at the lowest
concentrations, followed by voriconazole, caspofungin and
itraconazole. It has been shown previously that azole
antifungals were ineffective against C. albicans and Candida
dubliniensis biofilms, whereas echinocandins were the most
effective (Ramage et al., 2001a, b, c). Hawser et al. (2001)
reported that A. fumiga tus isolates w ere more susceptible to
echinocandins using conidial concentrations equivalent to
the CLSI M38-A method combined with XTT measure-
ments. However, in this study, with an increased cell
biomass, caspofungin was ineffective. The inability to
successfully treat IA patients with caspofungin has been
reported elsewhere, with only 41 % responding during
salvage therapy (Maertens et al., 2004). We note, however,
that both vori conazole and AmpB had the ability to reduce
cellular viability by over 50 % at relatively low concentra-
tions (Table 1). Therefore, although total death was not
achieved, there was a certain degree of efficacy against these
tenacious multicellular structures, which has been reported
previously for other in vitro fungal biofilms (Ramage et al.,
2002c). This may be an important observation when these
agents are used empirically, as they may have a role in
preventing the establishment of a mature biofilm. More
work will be required to elucidate this.
How do these results relate to clinical practice? A recent
review of empirical antifungal therapy in neutropenic
patients compared 13 studies. The success rates of the
treatments reported showed variations from 31 to 86 %
(Martino & Viscoli, 2006). This variability in outcome may
be as much to do with the patient population, the weakness
of the indications for treatment and the consequent
difficulty in establishing objective and reproducible end
points for comparisons as the effectiveness of the antifungal
agents. The time that treatment is started in relation to the
development of the mature biofilm may also impact on the
outcome. The optimal time for starting antifungal therapy in
neutropenic patients remains undetermined, although most
experts recommend waiting until day 5 or 7 of persistent
fever (Bennett et al., 2003). It may be more prudent to start
treatment before the multicellular structure has established,
but this will require further in vitro studies. In CF, there is
recognition that A. fumigatus in sputum cultures in the
absence of ABPA may be a pathogen that can directly cause
respiratory exacerbations. Antifungal therapy should be
considered when deteriorating respiratory function is not
responding to antibacterial therapy. Treatment with anti-
fungal agents has been evaluated in these patients and an
improvement in clinical condition observed (Shoseyov et al.,
2006). Nevertheless, more studies are required before the
effectiveness of antifungal agents in this group of patients
can be evaluated fully.
Overall, t his study demonstrated that A. fumigatus grows as
a complex, multicellular biofilm and that the co ncentration
of antifungal drug required for the effective treatment of
these biofilm-related infections is distinct from assessment
by the standard CLSI M38-A assay. This standard method
of susceptibility testing does not give a true evaluation of
the susceptibility of the disease-causing organism to a given
antifungal agent. This may have implications for the
diagnosis and management of patients with both invasive
and non-invasive infections with this organism. Future
studies to examine changes in gene expression of these
biofilm-associated organisms may provide ways of eluci-
dating new therapeutic options for con trolling A. fumigatus
biofilms in immunocompromised individuals.
ACKNOWLEDGEMENTS
We would like to thank Helen Kennedy for supplying the A. fumigatus
strains from the Royal Hospital for Sick Children (Yorkhill Division,
NHS Glasgow, UK). This study was supported by an educational
grant from Pfizer.
REFERENCES
Antachopoulos, C., Meletiadis, J., Roilides, E., Sein, T. & Walsh, T. J.
(2006). Rapid susceptibility testing of medically important zygomy-
cetes by XTT assay. J Clin Microbiol 44, 553–560.
Beauvais, A., Schmidt, C., Guadagnini, S., Roux, P., Perret, E., Henry, C.,
Paris, S., Mallet, A., Pre
´
vost, M.-C. & Latge
´
,J.P.(2007).An extracellular
matrix glues together the aerial-grown hyphae of Aspergillus fumigatus.
Cell Microbiol 9, 1588–1600.
Bennett, J. E., Powers, J., Walsh, T., Viscoli, C., de Pauw, B.,
Dismukes, W., Galgiani, J., Glauser, M., Herbrecht, R. & other
authors (2003).
Forum report: issues in clinical trials of empirical
antifungal therapy in treating febrile neutropenic patients. Clin Infect
Dis 36 , S117–S122.
Brakhage, A. A. (2005). Systemic fungal infections caused by
Aspergillus species: epidemiology, infection process and virulence
determinants. Curr Drug Targets 6, 875–886.
Chandra, J., Kuhn, D. M., Mukherjee, P. K., Hoyer, L. L., McCormick, T.
& Ghannoum, M. A. (2001).
Biofilm formation by the fungal pathogen
Aspergillus fumigatus biofilms
http://jmm.sgmjournals.org 1211

Candida albicans: development, architecture, and drug resistance. J
Bacteriol 183, 5385–5394.
Christensen, G. D., Simpson, W. A., Younger, J. J., Baddour, L. M.,
Barrett, F. F., Melton, D. M. & Beachey, E. H. (1985).
Adherence of
coagulase-negative staphylococci to plastic tissue culture plates: a
quantitative model for the adherence of staphylococci to medical
devices. J Clin Microbiol 22, 996–1006.
Cimon, B., Symoens, F., Zouhair, R., Chabasse, D., Nolard, N.,
Defontaine, A. & Bouchara, J. P. (2001). Molecular epidemiology of
airway colonisation by Aspergillus fumigatus in cystic fibrosis patients.
J Med Microbiol 50, 367–374.
Costerton, J. W., Lewandowski, Z., Caldwell, D. E., Korber, D. R. &
Lappin-Scott, H. M. (1995). Microbial biofilms. Annu Rev Microbiol
49, 711–745.
de Almeida, M. B., Bussamra, M. H. & Rodrigues, J. C. (2006).
Allergic bronchopulmonary aspergillosis in paediatric cystic fibrosis
patients. Paediatr Respir Rev 7, 67–72.
Denning, D. W., Marinus, A., Cohen, J., Spence, D., Herbrecht, R.,
Pagano, L., Kibbler, C., Kcrmery, V., Offner, F. & other authors
(1998).
An EORTC multicentre prospective survey of invasive
aspergillosis in haematological patients: diagnosis and therapeutic
outcome. EORTC Invasive Fungal Infections Cooperative Group. J
Infect 37, 173–180.
Ellis, M., Richardson, M. & de Pauw, B. (2000). Epidemiology. Hosp
Med 61, 605–609.
Filler, S. G. & Sheppard, D. C. (2006). Fungal invasion of normally
non-phagocytic host cells. PLoS Pathogens 2, e129.
Hawser, S. P., Norris, H., Jessup, C. J. & Ghannoum, M. A. (1998).
Comparison of a 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-
[
(phenyl-
amino)carbonyl
]
-2H-tetrazolium hydroxide (XTT) colorimetric
method with the standardized National Committee for Clinical
Laboratory Standards method of testing clinical yeast isolates for
susceptibility to antifungal agents. J Clin Microbiol 36, 1450–1452.
Hawser, S. P., Jessup, C., Vitullo, J. & Ghannoum, M. A. (2001).
Utility of 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-
[
(phenyl-ami-
no)carbonyl
]
-2H-tetrazolium hydroxide (XTT) and minimum effec-
tive concentration assays in the determination of antifungal
susceptibility of Aspergillus fumigatus to the lipopeptide class
compounds. J Clin Microbiol 39, 2738–2741.
Herbrecht, R., Denning, D. W., Patterson, T. F., Bennett, J. E., Greene,
R. E., Oestmann, J. W., Kern, W. V., Marr, K. A., Ribaud, P. & other
authors (2002).
Voriconazole versus amphotericin B for primary
therapy of invasive aspergillosis. N Engl J Med 347, 408–415.
Hope, W. W., Walsh, T. J. & Denning, D. W. (2005). Laboratory
diagnosis of invasive aspergillosis. Lancet Infect Dis 5, 609–622.
Kamai, Y., Chiang, L. Y., Lopes Bezerra, L. M., Doedt, T., Lossinsky,
A. S., Sheppard, D. C. & Filler, S. G. (2006).
Interactions of Aspergillus
fumigatus with vascular endothelial cells. Med Mycol 44 (Suppl. 1),
S115–S117.
Latge, J. P. (2001). The pathobiology of Aspergillus fumigatus. Trends
Microbiol 9, 382–389.
Maertens, J., Raad, I., Petrikkos, G., Boogaerts, M., Selleslag, D.,
Petersen, F. B., Sable, C. A., Kartsonis, N. A., Ngai, A. & other
authors (2004). Efficacy and safety of caspofungin for treatment of
invasive aspergillosis in patients refractory to or intolerant of
conventional antifungal therapy. Clin Infect Dis 39, 1563–1571.
Martino, R. & Viscoli, C. (2006). Empirical antifungal therapy in
patients with neutropenia and persistent or recurrent fever of
unknown origin. Br J Haematol 132, 138–154.
Meletiadis, J., Meis, J. F., Mouton, J. W. & Verweij, P. E. (2001a).
Analysis of growth characteristics of filamentous fungi in different
nutrient media. J Clin Microbiol 39, 478–484.
Meletiadis, J., Mouton, J. W., Meis, J. F., Bouman, B. A., Donnelly, J. P.
& Verweij, P. E. (2001b). Colorimetric assay for antifungal suscept-
ibility testing of Aspergillus species. J Clin Microbiol 39, 3402–3408.
Meletiadis, J., Mouton, J. W., Meis, J. F., Bouman, B. A., Donnelly, P. J.
& Verweij, P. E. (2001c).
Comparison of spectrophotometric and
visual readings of NCCLS method and evaluation of a colorimetric
method based on reduction of a soluble tetrazolium salt, 2,3-bis
[
2-
methoxy-4-nitro-5-
[
(sulfenylamino) carbonyl
]
-2H-tetrazolium-
hydroxide
]
, for antifungal susceptibility testing of Aspergillus species.
J Clin Microbiol 39, 4256–4263.
Mensi, M., Salgarello, S., Pinsi, G. & Piccioni, M. (2004). Mycetoma of
the maxillary sinus: endodontic and microbiological correlations.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod 98, 119–123.
NCCLS (2002). Reference Method for Broth Microdilution Antifungal
Susceptibility Testing of Filamentous Fungi. NCCLS document M38-A.
Pennsylvania, PA: National Committee for Clinical Laboratory
Standards.
O’Toole, G. A. & Kolter, R. (1998). Initiation of biofilm formation in
Pseudomonas fluorescens WCS365 proceeds via multiple, convergent
signalling pathways: a genetic analysis. Mol Microbiol 28, 449–461.
Ramage, G., Vande Walle, K., Wickes, B. L. & Lopez-Ribot, J. L.
(2001a). Standardized method for in vitro antifungal susceptibility
testing of Candida albicans biofilms. Antimicrob Agents Chemother 45,
2475–2479.
Ramage, G., Vande Walle, K., Wickes, B. L. & Lopez-Ribot, J. L.
(2001b).
Biofilm formation by Candida dubliniensis. J Clin Microbiol
39, 3234–3240.
Ramage, G., Vandewalle, K., Wickes, B. L. & Lopez-Ribot, J. L.
(2001c). Characteristics of biofilm formation by Candida albicans.
Rev Iberoam Micol 18, 163–170.
Ramage, G., Bachmann, S., Patterson, T. F., Wickes, B. L. & Lopez-
Ribot, J. L. (2002a). Investigation of multidrug efflux pumps in
relation to fluconazole resistance in Candida albicans biofilms. J
Antimicrob Chemother 49, 973–980.
Ramage, G., Saville, S. P., Wickes, B. L. & Lopez-Ribot, J. L. (2002b).
Inhibition of Candida albicans biofilm formation by farnesol, a
quorum-sensing molecule. Appl Environ Microbiol 68, 5459–5463.
Ramage, G., VandeWalle, K., Bachmann, S. P., Wickes, B. L. &
Lopez-Ribot, J. L. (2002c).
In vitro pharmacodynamic properties of
three antifungal agents against preformed Candida albicans biofilms
determined by time-kill studies. Antimicrob Agents Chemother 46,
3634–3636.
Ramage, G., Saville, S. P., Thomas, D. P. & Lopez-Ribot, J. L. (2005).
Candida biofilms: an update. Eukaryot Cell 4, 633–638.
Rivera, A., Hohl, T. & Pamer, E. G. (2006). Immune responses to
Aspergillus fumigatus infections. Biol Blood Marrow Transplant 12,47–49.
Shibuya, K., Ando, T., Hasegawa, C., Wakayama, M., Hamatani, S.,
Hatori, T., Nagayama, T. & Nonaka, H. (2004).
Pathophysiology of
pulmonary aspergillosis. J Infect Chemother 10, 138–145.
Shoseyov, D., Brownlee, K. G., Conway, S. P. & Kerem, E. (2006).
Aspergillus bronchitis in cystic fibrosis. Chest 130, 222–226.
Singh, N. (2005). Invasive aspergillosis in organ transplant recipients:
new issues in epidemiologic characteristics, diagnosis, and manage-
ment. Med Mycol 43 (Suppl. 1), S267–S270.
Tunney, M. M., Ramage, G., Field, T. R., Moriarty, T. F. & Storey, D. G.
(2004). Rapid colorimetric assay for antimicrobial susceptibility
testing of Pseudomonas aeruginosa. Antimicrob Agents Chemother 48,
1879–1881.
Villena, G. K. & Gutierrez-Correa, M. (2006). Production of cellulase
by Aspergillus niger biofilms developed on polyester cloth. Lett Appl
Microbiol 43, 262–268.
E. Mowat and others
1212 Journal of Medical Microbiology 56