Antimicrobial effect of fluoroquinolones for the eradication of nontypeable Haemophilus influenzae isolates within biofilms.
ABSTRACT Biofilms can be defined as communities of microorganisms attached to a surface. Those bacterial biofilms cause serious problems, such as antibiotic resistance and medical device-related infections. Nontypeable Haemophilus influenzae (NTHi) is an important pathogen in respiratory infections, as it forms biofilms both in vitro and in vivo such as human middle ear. Recent reports indicate that otitis media, paranasal sinusitis and lower respiratory tract infections caused by Haemophilus influenzae have become more difficult to treat with oral antibiotic therapy. However, there has been no attention given to antibiotic eradication of NTHi biofilm. To investigate the antimicrobial effect of various antibiotics against NTHi biofilm formation, we conducted the following comparative study using both beta-lactamase-negative ampicillin (AMP)-susceptible (BLNAS) and AMP-resistant (BLNAR) NTHi strains. In a microtiter biofilm assay, both levofloxacin and gatifloxacin, of the fluoroquinolone antibiotic group, significantly inhibited biofilm formation by BLNAS and BLNAR NTHi in a dose-dependent fashion compared to ampicillin of the penicillin antibiotic group, cefotaxime of the cephalosporin antibiotic group, and both erythromycin and clarithromycin of the macrolide antibiotic group. Furthermore, in flow cell chamber studies, confocal laser scanning microscopy counted survival bacteria in mature biofilm had been treated with gatifloxacin, ampicillin, cefotaxime and erythromycin. Only gatifloxacin completely killed the BLNAR NTHi isolates within biofilms without regard to the thickness of biofilm formation. The results of this study suggest that fluoroquinolones potentially have a role in therapy against diseases caused by both BLNAS and BLNAR NTHi isolates within biofilms.
NAOSITE: Nagasaki University's Academic Output SITENAOSITE: Nagasaki University's Academic Output SITE
Antimicrobial Effect of Fluoroquinolones for the Eradication of
Nontypeable Haemophilus Influenzae Isolates within Biofilms
Kaji, Chiharu; Watanabe, Kiwao; Apicella, Michael A.;
Tohoku journal of experimental medicine, 214(2), pp.121-128;
Copyright (c) 2008 Tohoku University Medical Press
Antibiotic Eradication of H. Infl uenzae Biofi lms
Tohoku J. Exp. Med., 2008, 214, 121-128
Received December 20, 2007; revision accepted for publication January 8, 2008.
Correspondence: Dr. Hiroshi Watanabe, Division of Infectious Diseases, Department of Infectious Medicine,
Kurume University School of Medicine, 67 Asahi-machi, Kurume, Fukuoka 830-0011, Japan.
Antimicrobial Effect of Fluoroquinolones for the
Eradication of Nontypeable Haemophilus Infl uenzae
Isolates within Biofi lms
CHIHARU KAJI,1 KIWAO WATANABE,1 MICHAEL A. APICELLA2 and HIROSHI WATANABE3
1Department of Internal Medicine, Institute of Tropical Medicine, Nagasaki University, Nagasaki,
2Carver College of Medicine, Department of Microbiology, The University of Iowa, Iowa City, IA,
3Division of Infectious Diseases, Department of Infectious Medicine, Kurume University School
of Medicine, Fukuoka, Japan
Biofi lms can be defi ned as communities of microorganisms attached to a surface. Those
bacterial biofi lms cause serious problems, such as antibiotic resistance and medical device-
related infections. Nontypeable Haemophilus infl uenzae (NTHi) is an important pathogen
in respiratory infections, as it forms biofi lms both in vitro and in vivo such as human mid-
dle ear. Recent reports indicate that otitis media, paranasal sinusitis and lower respiratory
tract infections caused by Haemophilus infl uenzae have become more diffi cult to treat with
oral antibiotic therapy. However, there has been no attention given to antibiotic eradica-
tion of NTHi biofi lm. To investigate the antimicrobial effect of various antibiotics against
NTHi biofi lm formation, we conducted the following comparative study using both β-lac-
tamase-negative ampicillin (AMP)-susceptible (BLNAS) and AMP-resistant (BLNAR)
NTHi strains. In a microtiter biofi lm assay, both levofl oxacin and gatifl oxacin, of the fl uo-
roquinolone antibiotic group, signifi cantly inhibited biofi lm formation by BLNAS and
BLNAR NTHi in a dose-dependent fashion compared to ampicillin of the penicillin antibi-
otic group, cefotaxime of the cephalosporin antibiotic group, and both erythromycin and
clarithromycin of the macrolide antibiotic group. Furthermore, in fl ow cell chamber stud-
ies, confocal laser scanning microscopy counted survival bacteria in mature biofi lm had
been treated with gatifl oxacin, ampicillin, cefotaxime and erythromycin. Only gatifl oxacin
completely killed the BLNAR NTHi isolates within biofi lms without regard to the thick-
ness of biofi lm formation. The results of this study suggest that fl uoroquinolones poten-
tially have a role in therapy against diseases caused by both BLNAS and BLNAR NTHi
isolates within biofi lms. ──── H. infl uenzae; BLNAS; BLNAR; biofi lm; fl uoroquino-
lone; ceftriaxone; ampicillin; macrolide.
Tohoku J. Exp. Med., 2008, 214 (2), 121-128.
© 2008 Tohoku University Medical Press
C. Kaji et al.
otics on NTHi biofi lms.
MATERIALS AND METHODS
Bacterial strains and culture conditions
The H99 and H01 strains were clinical isolates
obtained from different patients with respiratory tract
infections. Before this experiment, microtiter biofi lm
assay was performed in each of approximately 20
BLNAS and 20 BLNAR strains obtained from different
patients. The H99 and H01 strains were selected,
because these strains had formed biofi lms equally well.
NTHi strains were reconstituted from frozen stocks and
propagated on brain heart infusion (BHI) agar, or medi-
um (Difco Laboratories, Detroit, MI, USA) supplement-
ed with 10 μg/ml of hemin (Sigma Chemical Co., St.
Louis, MO, USA) and 10 μg/ml of β-nicotinamide-
adenine nucleotide (β-NAD) (Sigma) at 37°C and 5%
CO2 (Anderson et al. 1972).
Serotyping and PCR
H99 and H01 strains were serotyped by slide agglu-
tination with antisera purchased from Difco Laboratories
(Detroit, MI, USA). PCR was carried out for H. infl uen-
zae isolates by using mixed primers including the follow-
ing (Wakunaga Pharmaceutical Co., Hiroshima): P6
primers to amplify the p6 gene, which encodes the P6
membrane protein specifi c for H. infl uenzae; TEM-1
primers to amplify a part of the β-lactamase (bla) TEM-1
gene; pbp3-1 primers to identify the same substitution as
the low-BLNAR strains in the fts I gene; and pbp 3-2
primers to identify the same substitution as the BLNAR
strains in the fts I gene (Hasegawa et al. 2004).
Antimicrobial susceptibility test
Minimum Inhibitory Concentrations (MICs) were
determined using a broth microdilution method, accord-
ing to guidelines from the Clinical and Laboratory
Standards Institute (2007). Susceptibility of NTHi
strains to the following 6 antibiotics was evaluated:
ampicillin (AMP, Wako Pure Chemical Industries, Ltd.,
Osaka), cefotaxime (CTX, Wako), erythromycin (ERY,
Dainippon Pharmaceutical Co., Osaka), clarithromycin
(CLR, Abbott Japan Co., Tokyo), levofl oxacin (LVX,
Daiichi Sankyo Co., Ltd., Tokyo) and gatifl oxacin (GAT,
Kyorin Pharmaceutical Co., Tokyo). The antibiotics
evaluated were those commonly approved for the treat-
ment of bacterial infections from NTHi.
Nontypeable Haemophilus infl uenzae (H.
infl uenzae) is a gram-negative, pleomorphic bac-
terium that colonizes the human nasopharynx.
NTHi can cause a variety of infections, including
otitis media, sinusitis, conjunctivitis, bronchitis,
and pneumonia (Murphy and Apicella 1987;
Koyama et al. 2007). Recent reports state that
β-lactamase-negative ampicillin (AMP)-resistant
(BLNAR) strains have increased in some coun-
tries (Suzuki et al. 2003; Castanheira et al. 2006;
Sunakawa and Farrell 2007), although their global
prevalence remains low (Hoban and Felminghan
2002). Resistance in BLNAR strains results from
mutations in the ftsI gene-encoding, penicillin-
binding protein (pbp) 3, which mediates septal
peptidoglycan synthesis (Hasegawa et al. 2004).
Recent reports indicate that otitis media, paranasal
sinusitis and lower respiratory tract infection
caused by H. infl uenzae have become more
diffi cult to treat with oral antibiotics therapy
(Kaczmarek et al. 2004).
Biofi lm is a structured community of bacte-
rial cells enveloped in a self-produced, extra-
polymeric matrix that is adherent to surfaces
(Costerton et al. 1999; O’Toole et al. 2000).
Bacterial biofi lms are recognized as important
causes of a variety of human infections, including
infections of prosthetic devices, endocarditis, den-
tal caries, pneumonia in cystic fi brosis, prostatitis
and others (Singh et al. 2000; Donlan et al. 2001).
It is well known that bacteria within biofi lms are
more resistant to antibiotic therapy than are plank-
tonic organisms (Slinger et al. 2006). In order to
develop strategies for the treatment of infections
caused by bacteria in biofi lms, including otitis
media, it is important to elucidate the characteris-
tics of bacterial pathogens within biofi lms.
Recent reports state that NTHi forms bio-
fi lms both in vitro and in vivo, suggesting that
biofi lm formation in vivo might play an important
role in the pathogenesis and chronicity of otitis
media (Greiner et al. 2004; Jurcisek et al. 2005;
Hall-Stoodley et al. 2006). Very little information
is available concerning the effectiveness of vari-
ous antimicrobials for the eradication of NTHi
within biofi lms. The aim of our study is to evalu-
ate the antimicrobial effect of a variety of antibi-
Antibiotic Eradication of H. Infl uenzae Biofi lms
Microtiter biofi lm assay
Biofi lm formation in the H99 and H01 strains were
assessed using a 96-well microplate, as previously
described (Murphy and Kirkham 2002; Greiner et al.
2004). After being cultured for 48 hrs, the biofi lms were
exposed for 1 hr to differing concentrations (0.1 x MIC,
1 x MIC and 10 x MIC) of each antibiotics (AMP, CTX,
ERY, CLR, LVX and GAT) solution every 12 hrs for 2
days (totally 4 doses). The evaluation of biofi lm was
performed immediately after the fi nal exposure to the
antibiotics solution. The culture medium containing
planktonic cells was stained with 1% crystal violet at
room temperature. After rinsing with water three times,
the dye bound to the biofi lm was extracted with 230 μl of
95% ethanol for 15 min. The extracted dye was quanti-
fi ed by measuring the absorbance at 600 nm (OD600)
with a microplate reader. The strains were tested in
quadruplicate for each experiment and the results were
reported as three different experiments.
Biofi lm growth in continuous fl ow cell chamber
Biofi lms were formed in continuous fl ow cell cham-
bers as described by Davies et al. (1998). The BLNAR
H01 strain was grown in RPMI 1640 medium (Gibco
BRI, Grand Island, NY, USA) supplemented with 1
μg/ml of protoporphyrin IX which is hemin without iron,
0.1 mg/ml of hypoxanthine, 0.1 mg/ml of uracil, 10
μg/ml of β-NAD, 0.8 mM of sodium pyruvate and 100
μM sialic acid, or supplemented BHI medium. The cul-
ture was diluted to an optical density (OD) of 0.15 at 600
nm and supplemented with RPMI 1640 medium. The
cell suspensions of 1 × 108 colony-forming unit (cfu) /ml
was used to inoculate a 37 × 5 × 5 mm fl ow cell cham-
ber, fi lled with the prepared inoculum and incubated at
37°C for 1 hr to allow adherence to the glass coverslips
(Seib et al. 2007). The chambers were then incubated for
48 hrs under 125 μl/min fl ow in supplemented fresh
RPMI 1640 or BHI medium diluted 1:10 with sterile
phosphate-buffered saline (PBS). After being cultured
for 48 hrs, the biofi lms were exposed for 1 hr to differing
concentrations (0.1 x MIC, 1 x MIC and 10 x MIC) of
each antibiotics (AMP, CTX, ERY and GAT) solution
every 12 hrs for 2 days (totally 4 doses). After each
exposure, the medium-containing antibiotic was replaced
with fresh medium containing no antibiotic.
Approximately 5 hrs after the fi nal exposure to antibiotic
solution, the biofi lms were tested using a Dead/Live
BacLight viability stain kit (Molecular Probes, Eugene,
OR, USA), following the manufacturer’s protocol, by
counting the colonies.
Colony counting NTHi isolates within biofi lms
To determine the antimicrobial effect of each antibi-
otics against NTHi isolates within biofi lms, colony
counting was performed after 48 hrs. To collect bacteria
from the culture, the biofi lms were removed from the
fl ow chambers by fl ushing twice with 1 ml of sterile PBS
at the end of each experiment. These washes were com-
bined and aliquots (0.1 ml) of the suspension that had
been subjected to serial 10-fold dilutions. These aliquots
were sub-cultured onto supplemented BHI agars incubat-
ed at 37°C. The resultant colonies were counted to deter-
mine surviving colony-forming units (n = 3) at 48 hrs.
Quantifi cation of biofi lm structures
To evaluate the viability of biofi lm structure after
exposure to antibiotics, the biofi lm formations were anal-
ysed by COMSTAT (BioCentrum-DTU, Lungby,
Denmark) (Heydorn et al. 2000; Seib et al. 2007).
COMSTAT is a program for quantifi cation of three-
dimensional biofi lm structures. It analyzes stacks of
images acquired with confocal laser scanning microsco-
py (CLSM). First, biofi lms were stained using a Live/
Dead BacLight bacterial viability stain kit (Molecular
Probs, Eugene, OR, USA). This permits visualization of
live and dead bacteria within the biofi lms. Briefl y,
SYTO 9 and propidium iodide were mixed at a 1:1 ratio
in sterile PBS solution. The staining solution was intro-
duced into the chamber for 15 min at 37°C in the dark.
Biofi lm bacteria within the chamber were immediately
visualized with CLSM (Nicon DIGITAL ECLIPSE C1,
Nikon, Melville, NY, USA) using a modifi ed stage.
Three independent biofi lm experiments were performed
and at least four stacked images at 200 x magnifi cation
per experiment were obtained and average values were
used for statistical analyses. Each of the four stacked
images was selected from an apical surface area at ran-
dom in the chamber and examined. The threshold value
that best fi t all image stacks of a trial was chosen and
kept consistent for all stacks within the trial. Images
were acquired at 2.0 μm intervals down through the
biofi lm Z-stack (μm). Therefore, the number of images
in each stack varied according to the biomass and
average thickness of the biofi lm (Starner et al. 2006).
Quantifi cation of biofi lm biomass and average thickness
was calculated using a MatLab 5.3 (The MathWork, Inc.,
Natick, MA, USA), equipped with an Image Processing
Toolbox and COMSTAT analysis. The image stacks of
C. Kaji et al.
“without” and “with” following exposure to each con-
centration of the antibiotic medium groups were aver-
aged and compared.
Statistical tests and data analysis
All analyses of statistical signifi cance were per-
formed with one-tailed Student’s t-tests using Microsoft
Excel (Microsoft Corp., Redmond, WA, USA). P values
less than were considered statistically signifi cant. For
quantifi cation of biofi lm mass, each datapoint used was
the mean biomass of four randomly imaged areas from a
single biofi lm. Mean biomass values from different bio-
fi lm experiments were used for further statistical analyses
(Starner et al. 2006).
Characteristics of strains
The H99 and H01 strains used in this study
are listed in Table 1. The H99 and H01 strain
were both nontypeable as assayed by slide agglu-
tination with antisera. The MIC of H99 against
AMP was 0.06 mg/l and H99 did not have the
mutation of pbp 3-1 and pbp 3-2 by PCR, whereas
the MIC of H01 against AMP was 1.0 mg/l and
H01 had the mutation of pbp 3-1 and pbp 3-2 by
Fig. 1. Antimicrobial effect of 6 antibiotics against NTHi isolates within biofi lms between BLNAS (H99,
white bars) and BLNAR (H01, diagonally striped bars) strains by microtiter biofi lm assay.
1 x and 10 x MIC of cefotaxime, and 10 x MIC of clarithromycin signifi cantly inhibited only the
BLNAS NTHi biofi lms. By comparison, 1 x and 10 x MIC of levofl oxacin and gatifl oxacin inhibit-
ed both the BLNAS and BLNAR NTHi biofi lms. Results are given as the average of 3 replicates
with each performed in quadruplicate ± S.D.
p < 0.001 as compared with the same strain without exposure to antibiotics.
T 1. MICs of antibiotics and resistant gene in this study.
MIC (mg/L) Resistant gene by PCRg
TEM-1 pbp3-1 pbp3-2
aAMP, ampicillin. bCTX, cefotaxime. cERY, erythromycin. dCLR, clarithromycin. eLVX,
levofl oxacin. fGAT, Gatifl oxacin.
β-lactamase gene; TEM-1: -, None.
Penicillin binding protein genes; pbp3-1 and pbp3-2: +, altered; -, not altered.
fGAT, Gatifl oxacin.
fgTEM-1, type β-lactamase gene; TEM-1: -, None.