1. No category

Microbiology (2003), 149, 2397–2405
DOI 10.1099/mic.0.26434-0
Klebsiella pneumoniae MrkD-mediated biofilm
formation on extracellular matrix- and collagencoated surfaces
Jennifer Jagnow and Steven Clegg
Correspondence
Department of Microbiology, College of Medicine, University of Iowa, Iowa City, IA 52242, USA
Steven Clegg
steven-clegg@uiowa.edu
Received 22 April 2003
Revised 9 June 2003
Accepted 12 June 2003
The type 3 fimbriae of Klebsiella pneumoniae are comprised of the major fimbrial subunit (MrkA) and
the adhesin (MrkD) that has previously been shown to mediate binding to collagen. The ability
of adhesive and non-adhesive derivatives of K. pneumoniae to form biofilms on collagen-coated
surfaces in continuous-flow chambers was investigated. Unlike biofilm formation on abiotic
plastic surfaces, the presence of the MrkD adhesin was necessary for growth on collagen-coated
surfaces. Fimbriate strains lacking the MrkD adhesin did not efficiently adhere to and grow on
these surfaces. Similarly, purified human extracellular matrix and the extracellular matrix formed
by human bronchial epithelial cells grown in vitro provided a suitable substrate for MrkD-mediated
biofilm formation, whereas direct binding to the respiratory cells was not observed. Type 3
fimbriae may therefore have two roles in the early stages of adherence and growth on in-dwelling
devices such as endotracheal tubes. The MrkA polypeptide could facilitate adsorption to abiotic
polymers of recently implanted devices and the MrkD adhesin could enable bacteria to adhere to
and grow on polymers coated with host-derived proteins.
INTRODUCTION
The role of bacterial fimbriae in the production and
formation of biofilms has been described for several species
(Costerton et al., 1987, 1999; Watnick & Kolter, 2000). Type
1 fimbriae and the associated FimH adhesin of Escherichia
coli have been shown to influence biofilm formation in vitro
on plastic surfaces in static batch culture conditions and in a
hydrodynamic environment, respectively (Pratt & Kolter,
1998; Schembri & Klemm, 2001). Similarly, the type IV pili
of Pseudomonas aeruginosa have been demonstrated to
influence biofilm formation by these bacteria (O’Toole &
Kolter,1998). Also, the pili produced by Vibrio cholerae have
been shown to play a role in biofilm production (Watnick &
Kolter, 1999). The development of biofilms by pathogenic
bacteria is believed to play an important role in facilitating
evasion of host defence mechanisms, communication between
bacterial cells and protection against antibiotic action.
Recent data indicate that the expression of some genes in
sessile, biofilm-producing bacteria may be different to that
of planktonic cells (Whitely et al., 2001). Consequently, the
production of bacterial biofilms appears to be dependent on
multiple genetic factors. Therefore, surface components of
the bacterial cell that increase the efficiency of biofilm
Abbreviations: ECM, extracellular matrix; HBE, human bronchial
epithelial.
A real-time video of biofilm formation by K. pneumoniae IA565 is
available as supplementary data with the online version of this paper at
http://mic.sgmjournals.org.
0002-6434 G 2003 SGM
formation are likely to play a significant role in the
establishment of infection by pathogens.
We have recently demonstrated that the type 3 fimbriae of
Klebsiella pneumoniae influence the development of biofilms
in plastic, continuous flow-through chambers (Langstraat
et al., 2001). The type 3 fimbriae are expressed by several
species of opportunistic pathogens and are characterized by
their ability to mediate agglutination, in vitro, of treated
erythrocytes (Clegg et al., 1994; Old & Adegbola, 1985; Old
et al., 1985). The MrkA polypeptide comprises the major
structural component of type 3 fimbriae and is polymerized
to form the fimbrial shaft. The MrkD protein is believed to
function as the type 3 fimbrial adhesin and mediates binding
to extracellular matrix (ECM) proteins such as collagen
molecules (Schurtz et al., 1994; Schurtz Sebghati et al., 1998;
Tarkkanen et al., 1990, 1992). The efficient development of
biofilms by K. pneumoniae on plastic surfaces was independent of the presence of the MrkD adhesin, but facilitated by
the presence of the fimbrial shaft on the bacterial surface.
The collagen-binding MrkD molecule appears to play little
role in the development of biofilms on these plastic surfaces,
whereas the MrkA shaft protein is an important contributing factor. Therefore, bacterial binding leading to colonization and biofilm formation on plastic devices such as
endotracheal tubes may be more efficient by type 3 fimbriate
bacteria. However, in-dwelling devices such as catheters and
tubes are coated over time, in situ, with host-derived material
(Donlan, 2001; Francois et al., 1998). Consequently, specific
receptor–ligand binding, such as the MrkD–collagen
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
Printed in Great Britain
2397
J. Jagnow and S. Clegg
interaction, could also play a role in biofilm development
during infection. Also, MrkD-mediated binding to extracellular matrices is likely to have a role in colonization of
damaged epithelial surfaces.
During the course of our studies investigating the role of
type 3 fimbriae in biofilm development on abiotic plastic
surfaces we observed that fimbriate bacteria were limited in
their ability to form biofilms on glass surfaces. Consequently,
coverslips coated with ECM or collagen should provide a
good surface to investigate the role of MrkA and MrkD
proteins of the type 3 fimbriae on biofilm formation in
continuous-flow chambers coated with these substances.
METHODS
Bacterial strains and plasmids. K. pneumoniae IA565 is a clinical
isolate expressing type 3 fimbriae (Fim+) that mediate adherence
(Adh+) to collagen and treated erythrocytes as described previously
(Hornick et al., 1995). K. pneumoniae IApc35 is a mrkD mutant of
strain IA565 that is phenotypically fimbriate (Fim+), but lacks adhesive ability (Adh2) and its construction and characterization has
been reported elsewhere (Schurtz Sebghati et al., 1998; Tarkkanen
et al., 1997). The non-fimbriate (Fim2), non-adhesive (Adh2)
strain, IADT3, is a mrkB mutant of K. pneumoniae IApc35 and its
construction has been described previously (Tarkkanen et al., 1997).
All strains were grown on GCAA medium for optimal expression of
type 3 fimbriae and cultures were incubated at 37 uC for 18–24 h.
K. pneumoniae IADT3(pFK12) possesses the plasmid carrying the
complete mrk gene cluster and is restored in type 3 fimbrial expression and adhesiveness (Allen et al., 1991; Gerlach & Clegg, 1988;
Langstraat et al., 2001). To monitor biofilm development, bacteria
expressing green fluorescent protein (GFP) were used and these
strains have been described elsewhere (Langstraat et al., 2001).
Detection of type 3 fimbriae. The ability of bacteria to produce
type 3 fimbriae was determined using monospecific antisera raised
against purified fimbriae as reported previously (Schurtz et al.,
1994). Haemagglutinating activity of K. pneumoniae strains was
determined using tanned erythrocytes according to the method of
Old et al. (1985). Bacterial binding to collagen was performed using
the ELISA described in detail previously (Schurtz Sebghati et al.,
1998). Purified human placenta collagen (types V and IV) and ECM
were purchased from BD Biosciences and coating concentrations
were determined by standard techniques (Korhonen et al., 1997;
Kukkonen et al., 1993; Tarkkanen et al., 1997).
Biofilm formation on uncoated glass surfaces. The once
flow-through continuous culture system, developed by Parsek &
Greenberg (1999), was used to assay biofilm development on glass
slides that are affixed to the top of the flow cell. Biofilm development on the glass slides incubated at 37 uC using GCAA broth
(diluted 1 : 50) as the medium was examined at 6, 24, 48 and 72 h
post-inoculation. Following incubation the glass coverslips were
gently removed from the flow cell, ensuring that the correct orientation of the coverslip was maintained, and the coverslips were placed
on top of a clean, sterile flow-through chamber. In this way fluorescence from bacteria growing on the plastic surfaces of the biofilm
chamber could be eliminated. Subsequently, imaging of the biofilm
was performed using a Bio-Rad MRC600 confocal microscope and
images are presented as composite sections through the x–y planes
of these sections as described in detail elsewhere (Langstraat et al.,
2001; Parsek & Greenberg, 1999). The complete area of the slides
2398
exposed to the biofilm chamber was examined by confocal microscopy and representative sections were analysed to calculate the
depth of biofilm formed.
Biofilm formation on collagen- and cellular-coated coverslips. To examine biofilm formation on treated slides the following
modifications to the procedure described above were performed.
Collagen-coated glass or plastic coverslips were prepared as described previously using optimal coating concentrations of collagen
solution or human ECM (Schurtz et al., 1994). Collagen coating
(0?005 mg ml21) was performed overnight at 4 uC in carbonate
buffer by incubating the flow cell filled with coating solution, and
before inoculation the cell was flushed with PBS to remove excess
unbound collagen. When using human ECM the coating concentration was 0?109 mg ml21.
In a separate series of experiments, semiconfluent monolayers of a
human bronchial epithelial (HBE) cell line were also used to coat the
coverslips. Chambers were inoculated with 26105 cells and incubated
under 5 % CO2 for 24 h at 37 uC in Dulbecco’s Modified Eagle Medium
supplemented with 10 % fetal bovine serum and 25 mg chloramphenicol ml21. Under these conditions the coverslips are coated with a
semiconfluent monolayer of HBE cells. Following bacterial inoculation
the flow rate of the chamber was set at 130 ml min21 and the tissue
culture fluid plus 10 % fetal bovine serum was used as culture medium
that was supplemented with appropriate antibiotics. For the biofilm
assays the HBE cells were stained with 1 mM cell tracker orange
(Molecular Probes) according to the manufacturer’s instructions.
Finally, to examine the ability of K. pneumoniae strains to form biofilms
on the ECM produced in vitro by respiratory cells, the HBE cells were
grown as a confluent monolayer on the slides of the biofilm chamber.
Inoculation of the chamber with the HBE cells was as described above
and the cells were grown for 48 h. Under these conditions the HBE cells
were observed to form a continuous and confluent monolayer when
examined by microscopy. Subsequently, the cells were removed from
the slides by treatment in situ with 25 mM EGTA as described in detail
elsewhere (Hornick et al., 1995). This treatment results in effective
removal of the HBE cells without damaging the matrix laid down by
these cells. The ability of GFP-labelled bacteria to form biofilms on this
matrix was determined as described above.
The imaging of the biofilms on coated coverslips was performed as
described above for untreated surfaces. The biofilm development was
monitored at 6, 24 and 48 h post-inoculation for the ECM- and
collagen-coated slides, and 24 h for the HBE-coated slides.
Bacterial adherence to HBE cells grown in vitro. Examination
of the binding of K. pneumoniae strains to HBE cells in vitro was
performed as described previously (Hornick et al., 1995).
RESULTS
Biofilm formation on glass coverslips
Previously we have demonstrated that the MrkD adhesin of
the type 3 fimbriae was not required for efficient biofilm
formation on abiotic plastic surfaces (Langstraat et al.,
2001). However, examination of glass slides used in the
flow cell indicated very little bacterial growth. Consequently,
K. pneumoniae IA565 (Fim+ Adh+) and its derivative
strains, IApc35 (Fim+ Adh2) and IADT3 (Fim2 Adh2),
were further analysed for their ability to form biofilms
on glass surfaces. For all three strains no bacterial growth on
the glass of the flow cell was observed even after 72 h
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
Microbiology 149
Klebsiella biofilm formation on matrices
incubation. This is in contrast to the extensive biofilm
formed on the plastic surfaces, previously described by our
group (Langstraat et al., 2001), by the fimbriate strains after
24 h incubation.
Bacterial growth on human ECM- and collagencoated surfaces
Since we have previously demonstrated that the MrkD
adhesin mediates binding in vitro to extracellular matrices
and collagen (Langstraat et al., 2001; Schurtz Sebghati et al.,
1998), we investigated the ability of K. pneumoniae strains to
grow on human ECM-coated surfaces in a continuous
culture system. Optimal coating concentrations of matrix
6h
and collagen on the coverslips have been determined previously (Schurtz et al., 1994; Tarkkanen et al., 1990) and
these conditions were used to coat the coverslips of the flowthrough chambers. Careful removal of the coverslips from
the chambers enabled us to examine growth of the GFPproducing bacteria on these surfaces in the absence of
fluorescence due to growth on any uncoated plastic of the
chambers. After 6 h incubation patchy areas of growth of
K. pneumoniae IA565 were observed on the ECM-coated
slides and the maximum depth of the biofilm formed in
these regions was approximately 12?5 mm (Fig. 1). No
significant growth of K. pneumoniae IApc35 or IADT3 was
observed on the slides after 6 h incubation.
24 h
48 h
IA565
IApc35
IADT3
Fig. 1. Production of a biofilm on the surfaces of human ECM-coated coverslips. Scanning confocal laser microscopy
composite analysis (x–y plane) of biofilm images formed by GFP-producing bacteria after 6, 24 and 48 h incubation. The
maximum depth (33 mm) of biofilm was exhibited by strain IA565 after 6–24 h incubation. The maximum depth of patchy
biofilm growth formed by strain IApc35 (middle row) was approximately 8 mm.
http://mic.sgmjournals.org
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
2399
J. Jagnow and S. Clegg
0h
16 h
24 h
28 h
20 h
32 h
Fig. 2. Time-course of biofilm formation by K. pneumoniae IA565 growing on human ECM over a 32 h time period at 37 6C.
The real-time video of this growth is available as supplementary data with the online version of this paper at
http://mic.sgmjournals.org.
Following incubation for 24 h the fimbriate and adhesive
strain, IA565, formed an extensive biofilm covering most of
the coverslip surface (Fig. 1) and the maximum biofilm
depth was determined to be 33 mm, almost three times
greater than that observed after 6 h incubation. After 24 h
incubation K. pneumoniae IApc35 (Fim+ Adh2) demonstrated no growth on the collagen as determined by the
lack of fluorescence associated with the treated coverslips
(Fig. 1). The non-fimbriate strain, IADT3, did not grow on
the collagen-treated surfaces after 24 h incubation.
At 48 h post-inoculation of the chambers, strain IA565
continued to exhibit dense growth on the collagen-coated
surfaces, although the maximum depth of the biofilm was
approximately 25?5 mm which is less than that observed at
24 h. However, even after 3 days incubation the fimbriate
but non-adhesive strain, IApc35, still exhibited only limited
growth and a similar reduction in the depth of biofilm was
observed. The non-fimbriate and non-adhesive strain was
not observed to grow on the collagen-coated slides under
any conditions of incubation.
Fig. 2 demonstrates the time-course of biofilm development
by K. pneumoniae IA565 on human ECM. Initial bacterial
growth was limited during the first 16 h of incubation.
2400
Subsequently, extensive growth occurred over the following
16 h with the formation of characteristic pillars of bacterial
growth. A real-time video of GFP-labelled K. pneumoniae
IA565 growing over a 32 h time period on human ECM is
available as supplementary data with the online version of
this paper at http://mic.sgmjournals.org.
Similar results were observed using coverslips coated with
collagen (Table 1). Only K. pneumoniae IA565 developed
extensive growth over 48 h incubation. Both K. pneumoniae
IApc35 and IADT3 exhibited little or no growth over the
complete area of the coverslips during the incubation period.
Growth of K. pneumoniae IADT3(pFK12) on
coated surfaces
K. pneumoniae IADT3 is a mrkB mutant of strain IA565
and has been shown by our group to be a constitutively
non-fimbriate strain (Tarkkanen et al., 1997). The plasmid
pFK12 contains the complete mrk gene cluster of strain
IA565 and transformants of the mrkB mutant are fully
restored to type 3 fimbrial expression (Tarkkanen et al.,
1997). Fig. 3 shows the ability of K. pneumoniae
IADT3(pFK12) to grow as a biofilm on chambers coated
with the ECM. Unlike the untransformed strain, type 3
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
Microbiology 149
Klebsiella biofilm formation on matrices
Table 1. Adherence and biofilm-forming properties of K. pneumoniae and its derivatives on
collagen-coated slides
++, Bacterial growth over the whole surface of the coverslip; +, patchy areas of bacterial growth
throughout the coverslip; 2, no bacteria or only isolated single bacteria observed. Numbers in parentheses indicate the mean depth (mm) of the biofilm.
Strain
6 h incubation
IA565
IApc35
IADT3
24 h incubation
IA565
IApc35
IADT3
48 h incubation
IA565
IApc35
IADT3
Relevant
phenotype
Biofilm on
collagen-coated surfaces
Biofilm in intercellular
spaces of HBE cells
MrkAD
MrkA
2
+ (19)
2
2
+
2
2
MrkAD
MrkA
2
++ (45)
2
2
++
2
2
MrkAD
MrkA
2
++ (36)
2
2
++
2
2
fimbriate transformants efficiently colonized and subsequently grew on the surfaces coated with the ECM.
Similarly, the pFK12 transformants were able to form
biofilms over a 24 h period on chambers coated with
collagen. However, the strongly fimbriate transformants did
not grow on untreated glass surfaces in the biofilm chambers. The ability of the K. pneumoniae strains to grow in the
presence of airway epithelial cells was determined. HBE cells
were grown to semiconfluency to detect biofilm formation
in the vicinity of the fluorescently stained epithelial cells.
The cell-tracker orange probe facilitated both the examination of the localization of the GFP-producing bacteria with
the HBE cells and confirmed that these cells were viable.
After 24 h incubation K. pneumoniae IA565 was observed to
form patchy areas of biofilm in the regions where the HBE
cells were clustered. The bacteria were localized to regions
adjacent to HBE cells at sites where these cells are most likely
to deposit the highest concentrations of an ECM. Strain
IApc35 (Fim+ Adh2) did not form areas of biofilm growth
and were occasionally observed as single bacteria among
the HBE cells and strain IADT3 (Fim2 Adh2) did not grow
on the coverslips even though viable clusters of HBE cells
were present. Fig. 4 shows the characteristic growth of
K. pneumoniae IA565 adjacent to the cultured HBE cells. As
reported previously by our group (Hornick et al., 1995),
conventional binding assays in which the bacteria are
incubated with HBE cells for 60 min indicated no extensive
localization of large numbers of bacteria to the margins of
the epithelial cells.
Biofilm formation on HBE-derived matrices in
the absence of cells
We have previously described the inability of fimbriate
strains of K. pneumoniae to adhere in vitro to confluent
http://mic.sgmjournals.org
monolayers of HBE cells (Hornick et al., 1995). Consistent
with this observation is the lack of biofilm formation on
HBE-coated coverslips in the chambers by any of the
Klebsiella strains (data not shown). However, if the HBE
cells were removed from the coverslips prior to infection
with bacteria, the following results were obtained. Both
K. pneumoniae IApc35 (Fim+ Adh2) and IADT3 (Fim2
Adh2) failed to grow over a 24 h incubation period with no
observable fluorescence on any region of the coverslips. In
contrast, K. pneumoniae IA565 formed an extensive biofilm
after 24 h incubation at 37 uC (Fig. 5). These images are
representative of biofilm formation over the complete
matrix-covered slides. These experiments were repeated three
times to ensure that the images generated are consistent with
bacterial growth at different locations on the slides. The red
fluorescence represents single HBE cells that were occasionally observed to remain attached to the slides after
treatment with EGTA. Transformation of K. pneumoniae
IApc35 with pFK12 to restore adhesiveness also resulted in
the ability of transformants to grow on the HBE-derived
matrix.
DISCUSSION
Many of the fimbrial types synthesized using the chaperoneusher pathway of assembly are comprised of at least two
structural protein components that play a direct role in
determining the architecture and receptor-binding specificity of the fimbriae (Hultgren & Normark, 1991; Hultgren
et al., 1991). The MrkA polypeptide of the K. pneumoniae
type 3 fimbriae is the major subunit, whereas the MrkD
adhesin mediates binding, in vitro, to collagen (Allen et al.,
1991; Clegg et al., 1994). Fimbriae from a variety of bacterial
species have been shown by numerous groups to facilitate
biofilm formation in vitro using several different detection
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
2401
J. Jagnow and S. Clegg
was not necessary for efficient biofilm formation on
untreated surfaces. During these investigations we observed
that fimbriate bacteria did not adhere to glass coverslips if
these coverslips were used to cover the plastic biofilm
chambers. Consequently, we have used treated coverslips in
the flow-through chambers to investigate whether the
MrkD/collagen interaction plays any role in facilitating
biofilm formation in vitro.
(a)
Coating of the coverslip surfaces with human ECM or
purified collagen allows for the efficient formation of
biofilms by fimbriate bacteria bearing the MrkD adhesin.
Over a prolonged period of incubation (48 h or longer) we
observed that the depth of the biofilm mass did decrease.
This may be due to removal of distal layers of bacteria by the
constant flushing of medium through the chambers once a
maximum depth is formed after approximately 24 h. Also,
we found that it was necessary to remove the coverslips
from the chambers to view biofilm formation since, as we
previously reported, GFP-producing fimbriate bacteria
grow and efficiently adhere to the poorly coated plastic
surfaces of the chamber and interfere with our observations
of the collagen-coated surface. The coverslips were removed
to ensure minimal disturbance of the biofilm that was
present, but biofilms formed after 48 h incubation may be
more fragile than those of earlier time points. It is clear,
however, that the fimbriate strain lacking the MrkD adhesin
did not form biofilms on the collagen as efficiently as the
wild-type strain.
(b)
Fig. 3. Growth of K. pneumoniae IADT3 (a) and IADT3(pFK12)
(b) on human ECM-coated surfaces after 24 h incubation.
Scanning confocal laser microscopy composite analysis of
biofilm image of an x–y series through a depth of 32 mm.
systems (O’Toole & Kolter, 1998; Parsek & Greenberg, 1999;
Schembri & Klemm, 2001). For E. coli type 1 fimbriae it has
been demonstrated that the FimH adhesin plays a role in
biofilm formation by fimbriate bacteria on abiotic surfaces
(Schembri & Klemm, 2001; Watnick & Kolter, 2000). Using
the type 3 fimbrial system of K. pneumoniae, we recently
reported that the MrkA subunit facilitated biofilm formation on untreated plastic surfaces of flow-through chambers
that enable biofilm morphology to be examined (Langstraat
et al., 2001). However, the MrkD adhesin did not appear to
play a role in this assay system as the fimbriate but nonadhesive strain, K. pneumoniae IApc35, could establish a
biofilm on abiotic plastic surfaces as rapidly as the wild-type
strain. Therefore, in the once flow-through chambers using
K. pneumoniae, the adhesin molecule of the type 3 fimbriae
2402
It is known that K. pneumoniae type 3 fimbriae do not
mediate adherence to confluent monolayers of HBE cells
(Hornick et al., 1992, 1995). In addition, we have demonstrated that these fimbriae facilitate binding to the basement
membrane of human lung tissue sections (Hornick et al.,
1992). Consequently, we decided to investigate whether the
MrkD adhesin would enable K. pneumoniae biofilms to be
formed on the ECM produced by HBE cells grown in vitro
on the coverslips of the chambers. In one series of experiments we ensured that the HBE cells were not grown to
confluency so that extracellular material produced by the
cells would be exposed on the glass surfaces. The location of
the biofilm produced by K. pneumoniae IA565 under these
conditions is consistent with the interaction of the fimbrial
adhesin with the ECM. Large areas of bacterial growth were
associated with clusters of HBE cells where the ECM is
most likely to be exposed to the bacteria. As previously
reported, adherence assays using HBE cells grown in a
similar way indicated poor direct binding to the HBE cells.
Consequently, we propose that the large areas of bacterial
growth around the cells in the chambers is due to MrkDmediated binding to the ECM by small numbers of fimbriate
bacteria. Subsequently, these bacteria grow and establish
biofilm formation in these regions. The non-adhesive strains
were not able to grow on the slides under these conditions.
Although we did not directly demonstrate that the MrkD
protein interacts with any specific type of collagen in the
ECM of the HBE cells, it is known that collagen is an integral
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
Microbiology 149
Klebsiella biofilm formation on matrices
(a)
(b)
(c)
Fig. 4. Scanning confocal laser microscopy of the growth of K. pneumoniae IA565 (a) and IApc35 (b) growing on and near
HBE cells after 24 h incubation at 37 6C. The HBE cells are labelled with cell-tracker orange. Large areas of the GFPproducing and fimbriate bacteria (IA565) are located in close proximity to the HBE cells and exhibit a yellow fluorescence. The
non-fimbriate strain (IApc35) does not adhere or grow around the HBE cells. Viable, uninfected HBE cells (c) are stained red
by the cell-tracker dye surrounded by a pale green fluorescent substrate.
component of the matrix synthesized by cells in vitro
(Fenwick et al., 2001; Hastie et al., 2002; Yurchenko &
O’Rear, 1994). However, the composition of the ECM
produced by the HBE cells has not been investigated in
detail. The ability of the MrkD adhesin to facilitate adherence and subsequent growth on ECMs was confirmed using
(a)
commercially available, purified human ECM. Extensive
and complete biofilm formation on this material was only
observed using the strain expressing the MrkD molecule.
As indicated above, we believe that the MrkD adhesin does
not mediate direct bacterial binding to the HBE cell
(b)
(c)
Fig. 5. Biofilm formation by MrkD-positive or -negative bacteria on an HBE-derived matrix following removal of HBE cells by
EGTA. (a) The confluent monolayer of HBE cells are efficiently removed by EGTA treatment leaving single isolated cells (red)
on the slide. (b) The fimbriate but non-adhesive strain, K. pneumoniae IApc35, does not grow following 24 h incubation.
(c) Extensive biofilm formation by K. pneumoniae IA565 on the matrix after 24 h at 37 6C.
http://mic.sgmjournals.org
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
2403
J. Jagnow and S. Clegg
membranes in the chambers. However, to confirm that the
MrkD adhesin binds to the HBE-generated matrix it was
possible to remove a confluent layer of HBE cells from the
coverslips to expose any material laid down by the HBE
cells. In this case extensive biofilm formation was only
observed with MrkD-positive bacteria. The ability of MrkDpossessing fimbriae to grow on the underlying matrix of
epithelial cells cultured in vitro strongly suggests that
disruption of an intact epithelial surface is necessary for
type 3 fimbriate bacteria to grow on host surfaces. This
is consistent with the clinical observations associated with
nosocomially acquired infections due to K. pneumoniae
(Craven et al., 1990; Duma, 1985; Riser & Noone, 1981).
The use of the once flow-through chambers coated with
matrix proteins or as a support for culturing cells from a
relevant host provides an excellent model system to investigate the interaction of K. pneumoniae with human cells and
tissues in a dynamic environment. In this respect it has
advantages over static and closed systems, but does differ
from the natural site of infection by presenting the bacteria
to host material in a fluid environment rather than at an
air/mucosal surface interface located in the respiratory
tract. However, the ability to investigate the stages of
adherence, colonization and growth on host-derived tissues
and matrices will facilitate a molecular analysis of gene
expression in K. pneumoniae.
In humans, susceptibility to infections of the airways by the
opportunist K. pneumoniae is most frequently associated
with predisposing factors. During secondary infections,
exfoliated and denuded epithelial surfaces may expose
collagenous receptors that enable MrkD-mediated adherence of K. pneumoniae. Subsequent growth at these sites in
the form of biofilms could enable the bacteria to avoid
efficient killing by alveolar macrophages. For patients with
in-dwelling endotracheal tubes the type 3 fimbriae may have
a dual role in the infective process. Immediately after
insertion of these devices the hydrophobic nature of the type
3 fimbrial shaft could facilitate attachment to the polymer
surfaces of these tubes with subsequent growth on the
device. Also, it has been demonstrated that, with time, these
tubes are coated, in situ, with host-derived material (Donlan,
2001; Francois et al., 1998). Therefore, a second role of the
type 3 fimbriae in these types of infection may involve a
specific receptor ligand interaction involving MrkDmediated adherence to matrix proteins covering the tubes.
Clegg, S., Korhonen, K. T., Hornick, B. D. & Tarkkanen, A.-M.
(1994). Type 3 fimbriae of the Enterobacteriaceae. In Fimbriae:
Aspects of Adhesion, Genetics, Biogenesis, and Vaccines, pp. 97–104.
Edited by Per Klemm. Boca Raton, FL: CRC Press.
Costerton, J. W., Cheng, K.-J., Geesey, G. G., Ladd, T. I., Nickel, J. C.,
Dasgupta, M. & Marrie, T. J. (1987). Bacterial biofilms in nature and
disease. Annu Rev Microbiol 41, 435–464.
Costerton, J. W., Stewart, P. S. & Greenberg, E. P. (1999). Bacterial
biofilms: a common cause of persistent infections. Science 284,
1318–1322.
Craven, D. E., Barber, T. W., Steger, K. A. & Montecalvo, M. A.
(1990). Nosocomial pneumonia in the 1990s: update of epidemio-
logy and risk factors. Semin Respir Infect 5, 157–172.
Donlan, R. (2001). Biofilm formation: a clinically relevant micro-
biological process. Clin Infect Dis 33, 1387–1392.
Duma, R. J. (1985). Gram-negative bacillary infections. Pathogenic
and pathophysiologic correlates. Am J Med 78, 154–164.
Fenwick, S. A., Curry, V., Clements, S., Hazleman, B. L. & Riley, G. P.
(2001). 96-well plate-based method for total collagen analysis of cell
cultures. Biotechniques 30, 1010–1014.
Francois, P., Vaudaux, P. & Lew, P. D. (1998). Role of plasma and
extracellular matrix proteins in the physiology of foreign body
infections. Ann Vas Surg 12, 34–40.
Gerlach, G.-F. & Clegg, S. (1988). Cloning and characterization of
the gene cluster encoding type 3 (MR/K) fimbriae of Klebsiella
pneumoniae. FEMS Microbiol Lett 49, 377–383.
Hastie, A. T., Kraft, W. K., Nyce, K. B., Zangrilli, J. G., Musani, A. I.,
Fish, J. E. & Peters, S. P. (2002). Asthmatic epithelial cell
proliferation and stimulation of collagen production. Am J Resp
Crit Care Med 165, 266–272.
Hornick, D. B., Allen, B. L., Horn, M. A. & Clegg, S. (1992).
Adherence to respiratory epithelia by recombinant Escherichia coli
expressing Klebsiella pneumoniae type 3 fimbrial gene products. Infect
Immun 60, 1577–1588.
Hornick, D. B., Thommandru, J., Smits, W. & Clegg, S. (1995).
Adherence properties of an mrkD-negative mutant of Klebsiella
pneumoniae. Infect Immun 63, 2026–2032.
Hultgren, S. J. & Normark, S. (1991). Biogenesis of the bacterial
pilus. Curr Opin Genet Dev 1, 313–318.
Hultgren, S. J., Normark, S. & Abraham, S. N. (1991). Chaperone-
assisted assembly and molecular architecture of adhesive pili. Annu
Rev Microbiol 45, 383–415.
Korhonen, T. K., Lahteenmaki, K., Kukkonen, M., Pouttu, R.,
Hynonen, U., Savolainen, K., Westerlund-Wikstrom, B. & Virkola, R.
(1997). Plasminogen receptors. Turning Salmonella and Escherichia
coli into proteolytic organisms. Adv Exp Med Biol 412, 185–192.
Kukkonen, M., Raunio, T., Virkola, R., Lahteenmaki, K., Makela, P. H.,
Klemm, P., Clegg, S. & Korhonen, T. K. (1993). Basement
membrane carbohydrate as a target for bacterial adhesion: binding of
type I fimbriae of Salmonella enterica and Escherichia coli to laminin.
Mol Microbiol 7, 229–237.
ACKNOWLEDGEMENTS
This work was supported by Public Health Service grants RO1 AI46473
and AI50011 (S. C.). We thank the personnel in the laboratory of E. P.
Greenberg for their helpful cooperation and assistance in developing
the biofilm assays used in these studies.
Langstraat, J., Bohse, M. & Clegg, S. (2001). The type 3 fimbrial
shaft (MrkA) of Klebsiella pneumoniae, but not the fimbrial
adhesin (MrkD), facilitates biofilm formation. Infect Immun 69,
5805–5812.
Old, D. C. & Adegbola, R. A. (1985). Antigenic relationships among
REFERENCES
type-3 fimbriae of Enterobacteriaceae revealed by immunoelectronmicroscopy. J Med Microbiol 20, 113–121.
Allen, B. L., Gerlach, G. F. & Clegg, S. (1991). Nucleotide sequence
Old, D. C., Tavendale, A. & Senior, B. W. (1985). A comparative
and functions of mrk determinants necessary for expression of type 3
fimbriae in Klebsiella pneumoniae. J Bacteriol 173, 916–920.
study of the type-3 fimbriae of Klebsiella species. J Med Microbiol 20,
203–214.
2404
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
Microbiology 149
Klebsiella biofilm formation on matrices
O’Toole, G. A. & Kolter, R. (1998). Flagellar and twitching motility
are necessary for Pseudomonas aeruginosa biofilm development. Mol
Microbiol 30, 295–304.
Tarkkanen, A. M., Allen, B. L., Westerlund, B., Holthofer, H.,
Kuusela, P., Risteli, L., Clegg, S. & Korhonen, T. K. (1990). Type V
Parsek, M. R. & Greenberg, E. P. (1999). Quorum sensing signals in
collagen as the target for type-3 fimbriae, enterobacterial adherence
organelles. Mol Microbiol 4, 1353–1361.
development of Pseudomonas aeruginosa biofilms. Methods Enzymol
310, 43–55.
Tarkkanen, A. M., Allen, B. L., Williams, P. H. & 7 other authors
(1992). Fimbriation, capsulation, and iron-scavenging systems of
Pratt, L. A. & Kolter, R. (1998). Genetic analysis of Escherichia coli
Klebsiella strains associated with human urinary tract infection. Infect
Immun 60, 1187–1192.
biofilm formation: roles of flagella, motility, chemotaxis, and type 1
pili. Mol Microbiol 30, 285–293.
Riser, E. & Noone, P. (1981). Klebsiella capsular type versus site of
isolation. J Clin Pathol 34, 552–555.
Schembri, M. A. & Klemm, P. (2001). Biofilm formation in a
hydrodynamic environment by novel fimH variants and ramifications for virulence. Infect Immun 69, 1322–1328.
Tarkkanen, A.-M., Virkola, R., Clegg, S. & Korhonen, T. K. (1997).
Binding of the type 3 fimbriae of Klebsiella pneumoniae to human
endothelial and urinary bladder cells. Infect Immun 65, 1546–1549.
Watnick, P. & Kolter, R. (2000). Biofilm, city of microbes. J Bacteriol
182, 2675–2679.
Schurtz, T. A., Hornick, D. B., Korhonen, T. K. & Clegg, S. (1994).
Watnick, P. I. & Kolter, R. (1999). Steps in the development of a
Vibrio cholerae El Tor biofilm. Mol Microbiol 34, 586–595.
The type 3 fimbrial adhesin gene (mrkD) of Klebsiella species is not
conserved among all fimbriate strains. Infect Immun 62, 4186–4191.
Whitely, M., Bangera, M. G., Bumgarner, R. E., Parsek, M. R., Teitzel,
G. M., Lory, S. & Greenberg, E. P. (2001). Gene expression in
Schurtz Sebghati, T. A., Korhonen, T. K., Hornick, D. B. & Clegg, S.
(1998). Characterization of the type 3 fimbrial adhesins of Klebsiella
Pseudomonas aeruginosa biofilms. Nature 413, 860–864.
strains. Infect Immun 66, 2887–2894.
Opin Cell Biol 6, 674–681.
http://mic.sgmjournals.org
Yurchenko, P. D. & O’Rear, J. J. (1994). Basal lamina assembly. Curr
Downloaded from www.sgmjournals.org by
IP: 93.91.26.97
On: Tue, 07 Jul 2015 07:57:29
2405