Screening of bacteria for their ability to adhere and to form biofilms on polystyrene
surfaces
In order to efficiently select strains able to adhere and to form biofilms, the whole
bacterial collection of 156 heterotrophic aerobic bacteria, isolated as described
in the “Methods†section, was first screened with a rapid method, based on crystal
violet staining of biofilms formed in 96-well microtiter plates (polystyrene surfaces).
The bacteria that efficiently formed biofilms onto this surface after a 2 h adhesion
step and 24 h of growth were identified by 16S rDNA sequencing. This screening revealed
that the biofilm formation ability was very variable according to the bacterial strains
(Fig. 1).
Fig. 1. Quantification of bacterial biofilm formation on polystyrene microtiter plates under
static conditions. Bacteria were isolated from intertidal mudflat biofilms (white
bars) and from corrosion product-microorganism composite biofilms developed on harbor
metallic structures (black bars). After 24 h of growth, single-species biofilms were
quantified with crystal violet and the ratios of OD
595
(cells grown in biofilm)/OD
600
(planktonic cells) were calculated. The ratios are represented on the y axis. Dotted bars: bacteria with a ratio OD
595
/OD
600
8. Bars represent means ± standard deviations for three replicates
Under our experimental conditions, out of 86 isolates from the intertidal mudflat
biofilms, 15 strains were able to form a biofilm with a ratio of cel1s grown in biofilm/planktonic
cells higher than 2 (Fig. 1). These biofilm-forming bacteria were distributed in 5 bacterial classes: Flavobacteriia
(27 %), Gammaproteobacteria (27 %), Alphaproteobacteria (20 %), Bacilli (20 %) and
Actinobacteria (6 %). The Flavobacterium sp. II2003 strain, with a ratio of 13, showed the best ability to form a biofilm
on polystyrene. Other bacteria displayed a strong ability to form a biofilm on polystyrene:
Postechiella sp. I4003, Roseobacter sp. I4016 and IV3009 and Shewanella sp. IV3014 showed a ratio higher than 8 (Fig. 1). The proportion of benthic bacterial strains able to form biofilms according to
the sampling time at low tide is presented in Fig. 2a. Bacteria forming a biofilm were found at all emersion times (from 2 to 4 h). However,
when less than 10 strains were isolated from a sample, no biofilm-forming bacterium
was detected in this sample, whatever the emersion time (Fig. 2a).
Fig. 2. Proportion of bacterial strains able to form biofilms on polystyrene microtiter plates
under static conditions. Results are presented according to the emersion time for
bacteria isolated from mudflat biofilms (a) or the immersion time for bacteria isolated from corrosion product-microorganism
composite biofilms (b). Mudflat sampling was performed three times at low tide during three days (D
1
, D
2
, and D
3
). White bars: number of strains tested. Black bars: number of forming-biofilm strains.
*: no forming-biofilm strain detected
Concerning the bacteria isolated from biofilms developed on corroded carbon steel
immersed in sea water, 10 strains among the 70 isolates were able to form biofilms
after 24 h (with a ratio of cells grown in biofilm/planktonic cells higher than 2)
(Fig. 1). These strains were affiliated to the same taxonomic groups as the benthic bacteria,
but the proportion of bacteria from each class varied: 50 % Alphaproteobacteria, 20
% Flavobacteriia, 10 % Gammaproteobacteria, 10 % Bacilli and 10 % Actinobacteria.
Under our experimental conditions, the ratios of cel1s grown in biofilm/planktonic
cells obtained for these bacteria were lower than for the benthic bacteria. The ratio
of 9, for the Roseovarius sp. VA014 strain, was the highest value obtained for bacteria isolated from corroded
structures (Fig. 1). Erythrobacter sp. IVA009 was also interesting with a ratio higher than 8 (Fig. 1). The results presented in Fig. 2b show that bacteria able to form a biofilm on polystyrene were found in all samples,
but the highest number was isolated from the steel immersed for 2 weeks.
In conclusion, this first screening allowed us to detect 15 benthic bacteria and 10
bacteria from corroded structures able to develop a biofilm in 96-well polystyrene
microplates.
Ability of the selected strains to adhere and to form biofilms under static conditions
on glass surfaces
The above screening method in polystyrene microplates was rapid and convenient to
detect the bacterial ability to form biofilms, but did not provide any structural
information on these biofilms. To get this kind of information and thus study more
accurately stable biofilms, the experiments had to be performed in dynamic conditions
with biofilm observation by confocal laser scanning microscopy. Such observations
required glass surfaces, and it was uncertain whether strains able to develop a biofilm
on polystyrene would also be able to do it on glass. Since the biofilm study in dynamic
conditions was labour intensive and time consuming, all bacteria that formed biofilms
on polystyrene were then screened for their ability to form a biofilm on glass surfaces,
first of all in static conditions. After a 2 h adhesion step and 24 h of growth, the
biofilm was stained with DAPI, and microscopic fluorescence observations were performed
to detect the strains behaviour on the glass surface. Among the previously selected
bacteria (15 benthic bacteria and 10 bacteria isolated from corroded structures),
only the Postechiella sp. I4003, Flavobacterium sp. II2003, Flaviramulus sp. II2004, Roseobacter sp. IV3009 and Shewanella sp. IV3014 benthic bacteria and the Roseobacter sp. IIIA017 and Roseovarius sp. VA014 strains from corroded structures were able to form biofilms under these
conditions (Fig. 3). Microscopic observations of their biofilms showed a high percentage of colonized
surfaces, from 42.5 % for Postechiella sp. I4003 to 76 % for Roseobacter sp. IV3009 (Fig. 3a). On the bases of the biofilm structures, two types of biofilms could be distinguished
(Fig. 3a). The Postechiella sp. I4003, Flavobacterium sp. II2003 and Roseobacter sp. IV3009 biofilms were very heterogeneous and contained large cell aggregates whereas
the Roseobacter IIIA017, Roseovarius sp. VA014, Flaviramulus sp. II2004 and Shewanella sp. IV3014 biofilms contained more evenly distributed cells. To classify the strains,
bacteria were gathered when they exhibited no significant biofilm thickness differences.
Thus, Flavobacterium sp. II2003, Roseobacter sp. IV3009, Shewanella sp. IV3014 and Roseovarius sp. VA014 were grouped. They built significantly thicker biofilms (40.2 ?m, 33.8
?m, 34 ?m and 31.1 ?m respectively, Fig. 3b). Roseobacter sp. IIIA017 and Flavobacterium sp. I4003 biofilms were significantly thinner than all other biofilms, with an average
of 16.4 ?m and 19.8 ?m respectively (Fig. 3b), and formed another group. Finally, Flaviramulus sp. II2004 biofilm exhibited an intermediate average thickness (24.2 ?m), significantly
different from all other strain biofilms (Fig. 3b).
Fig. 3. Fluorescence microscopic 3D reconstitutions and quantification of biofilms formed
on glass surfaces under static conditions. After 24 h of growth under static conditions,
biofilms were stained with DAPI. Microscopic 3D images were reconstituted (a), the average thicknesses of the biofilms were determined and the differences between
them were statistically tested (b). A, B, C, D, E: bacterial isolates from mudflat biofilms. F, G: bacterial isolates from corrosion product-microorganism composite biofilms. Scale
bar: 200 ?m. CP: percentage of colonized surface. These values are averages of data
from three independent experiments, with standard deviations lower than 10 % of each
value. *: p 0.05; **: p 0.01; ***: p 0.001. ns: not significant. Circles with the same color indicate bacteria with
no significant biofilm thickness differences. Blue circles: thickest biofilms. Red
circles: thinnest biofilms. Black circle: intermediate thickness
Study of the bacterial biofilm structures under dynamic conditions in flow cells
Through the previous steps, seven bacteria have been selected for their capability
to form a thick biofilm under static conditions on polystyrene as well as glass surfaces.
These bacteria were then studied under dynamic conditions, to further investigate
strains able to develop stable biofilms. Thus, bacterial biofilms were grown on glass
slides in three-channel flow cells and observed by confocal laser scanning microscopy
after staining with the Syto 61 fluorescent dye.
The seven strains were able to attach onto the glass slide during a 2 h adhesion step
in artificial seawater without flow, but the biofilms of three strains were not sufficiently
stable and only four strains (Flavobacterium sp. II2003, Roseobacter sp. IV3009, Shewanella sp. IV3014 and Roseovarius sp. VA014) were able to form biofilms after 24 h of growth under a continuous culture
medium flow. The microscopic observations of these strains are shown in Figs. 4 and 5. After the adhesion step, Flavobacterium sp. II2003, Shewanella sp. IV3014 and Roseovarius sp. VA014 began to form aggregates or microcolonies, whereas Roseobacter sp. IV3009 cells were more individually attached (Fig. 4). Roseovarius sp.VA014 cells were more filamentous. About 25 % of the glass surfaces were covered
for all strains, except for the Roseobacter sp. IV3009 strain, which exhibited a significantly (p??0.05) lower percentage of colonized surface (16 %, Fig. 4). After 24 h of biofilm growth, four different 3D architectures were observed (Fig. 5). Flavobacterium sp. II2003 biofilms harbouring numerous mushroom-like structures with a non-uniform distribution
are reminiscent of biofilms of the well-known Pseudomonas aeruginosa model 17], 18].
Fig. 4. Confocal laser scanning microscopy images of attached cells after 2 h of adhesion
on glass surfaces. Bacteria were allowed to attach into the flow cells during 2 h
in artificial seawater without flow. Syto 61 red was used to stain the attached cells.
a, b, c: bacterial isolates from mudflat biofilms. d: bacterial isolate from corrosion product-microorganism composite biofilms. Scale
bar: 47?m. CP: percentage of colonized surface. These values are averages of data
from three independent experiments, with standard deviations lower than 10 % of each
value
Fig. 5. Confocal laser scanning microscopy images of single-species biofilms formed after
24 h of growth on glass surfaces under dynamic conditions. Biofilms were grown on
glass surfaces in flow cells, at 22° C for 24 h, under a flow of Zobell medium. Bacteria
were stained with Syto 61 red. a, b, c: 3D views of biofilms of bacterial isolates from mudflat biofilms. d: 3D view of a biofilm of the bacterial isolate from corrosion product-microorganism
composite biofilms. Each image is representative of 10 observations. Scale bar: 67.3?m
The Flavobacterium sp. II2003 biofilms presented significantly higher maximal thicknesses compared to biofilms
of the three other strains, due to the mushroom-like structures, but their average
thicknesses were only significantly higher than those of the Shewanella sp. IV3014 biofilms (Fig. 6). Shewanella sp. IV3014 displayed hairy biofilms with horizontal fibres, whereas Roseovarius sp. VA014 developed heterogeneous and tousled biofilms with cell aggregates (Fig. 5). Roseobacter sp. IV3009 biofilms were quite homogeneous with a bacterial distribution covering
the entire surface (Fig. 5). The average and maximal thicknesses of Roseobacter sp. IV3009 biofilms were the same (10 ?m, Fig. 6), confirming the regular distribution of cells. No significant differences were observed
between Shewanella sp. IV3014, Roseovarius sp. VA014 and Roseobacter sp. IV3009 for the average and the maximal biofilm thicknesses (Fig. 6). Similarly, the biovolumes of all 4 biofilms were not significantly different from
each other.
Fig. 6. COMSTAT analyses of biofilms formed on glass surfaces, after 24 h of growth under
dynamic conditions. A, B, C: bacterial isolates from mudflat biofilms (white bars). D: bacterial isolate from corrosion product-microorganism composite biofilms (black
bars). Significant differences were only observed in isolate pairs A-B, A-B, A–D for maximal thickness, A–C for average thickness and are indicated by * (p 0.05) or ** (p 0.01) on the upper part of the Figure. In all the other cases, the differences
were not significant
In conclusion, only four strains, Flavobacterium sp. II2003, Roseobacter sp. IV3009, Shewanella sp. IV3014 from the mudflat and Roseovarius sp. VA014 from the steel structure, were able to develop a stable single-species
biofilm under dynamic conditions. Each biofilm had a specific structure. Interestingly,
these strains were the four bacteria that displayed the thickest biofilms (more than
30 ?m) on glass surfaces under static conditions (Fig. 3b).