Screening of biofilm production, morphotypes and phylotypes,
Large variations in biofilm formation in microtiter plates were observed between strains, with A595 values ranging from -0.027 to 3.333, and a mean A595 of 1.215 (Table 1). In total, 84.2 % of the strains were considered positive for biofilm production, i.e. displayed A595 values above three standard deviations of the negative controls (A595 cut-off = 0.085) (Table S1).
All strains but one produced biofilm on CR agar plates. The dominating biofilm morphotype was RDAR, which was displayed by 69.6 % of the strains (Table 1). The morphotypes BDAR and PDAR were displayed by 26.6 % and 3.2 %, respectively, whereas one strain (0.6 %) displayed the non-biofilm morphotype SAW. Mean biofilm production in microtiter plates was significantly higher in RDAR strains than in BDAR (p = 1.2*10-7) and PDAR strains (p = 0.0018) (Table 1). RDAR strains displayed both a higher percentage of positive biofilm producers, and a higher mean A595 of the biofilm positive strains (Table S1).
Table 1. Distribution of morphotypes in the CR agar plate assay, and biofilm production by the different morphotypes as indicated by A595 in the microtiter plate assay. Means with same letter are not statistically different (p > 0.05).
Morpho-type
|
No
|
% of all
|
Mean A595 ± SD
|
Min. A595
|
Max. A595
|
RDAR
|
110
|
69.6
|
1.505A ± 0.964
|
-0.002
|
3.333
|
BDAR
|
42
|
26.6
|
0.610B ± 0.764
|
-0.027
|
2.744
|
PDAR
|
5
|
3.2
|
0.143B ± 0.120
|
0.010
|
0.261
|
SAW
|
1
|
0.6
|
0.083
|
-
|
-
|
ALL
|
158
|
100.0
|
1.215 ± 1.000
|
-0.027
|
3.333
|
Min. = minimum, Max. = maximum, SD= standard deviation.
Phylotype B2 was the most common (47 %), followed by D (24 %), B1 (16 %) and A (13 %). There were no significant differences in mean biofilm production in microtiter plates between phylotypes in the total material (Table 2). However, significant differences between phylotypes were observed within morphotypes RDAR and BDAR. Furthermore, biofilm production was displayed by all BDAR strains with phylotypes A and B1, but only by 63.6 and 18.8 % of BDAR strains with the phylotypes B2 and D, respectively. Distributions of morphotypes varied between phylotypes (Figure S1). RDAR dominated within phylotypes A and B2 (76.2 % and 77.9 % of the strains, respectively), whereas a more equal distribution of RDAR and BDAR strains were observed within phylotypes B1 (56.0 % vs 44.0 %) and D (57.9 % vs 42.1 %).
Table 2. Distribution of phylotypes, and mean biofilm production by the different phylotypes as indicated by A595 in the microtiter plate assay, in total material and within morphotypes RDAR and BDAR. Means with same letter are not statistically different (p > 0.05).
|
TOTAL
|
RDAR
|
BDAR
|
Phylo-type
|
No
|
%
|
Mean A595 ± SD
|
No
|
%
|
Mean A595 ± SD
|
No
|
%
|
Mean A595
± SD
|
|
A
|
21
|
13.3
|
1.345 ± 1.114
|
16
|
14.5
|
1.503AC ± 1.158
|
4
|
9.5
|
1.029A ± 0.877
|
|
B1
|
25
|
15.8
|
1.157 ± 0.821
|
14
|
12.7
|
1.245A ± 0.896
|
11
|
26.2
|
1.046A ± 0.741
|
|
B2
|
74
|
46.8
|
1.129 ± 0.862
|
58
|
52.7
|
1.341A ± 0.823
|
11
|
26.2
|
0.458D ± 0.559
|
|
D
|
38
|
24.1
|
1.350 ± 1.274
|
22
|
20.0
|
2.105BC ± 1.017
|
16
|
38.1
|
0.311D ± 0.753
|
|
SD = standard deviation.
When comparing strains originating from chicken caecum and retail chicken meat, there were no significant differences in mean biofilm formation (Table 3) or the percentage of biofilm producers in the microtiter plate assay (89,4 % and 83.6 %, respectively, chi square test p = 0.31). Neither were there any differences in distribution of biofilm morphotypes or phylotypes (Tables 3 and S2). However, within the most common morphotype RDAR, caecal strains displayed a higher mean biofilm A595 than those from meat (Table 3). Also within phylotype D, mean biofilm production was significantly higher in caecal samples than in meat samples (Table S2).
Table 3. Comparison of biofilm production by isolates from chicken caecal and retail meat samples as indicated by A595 in the microtiter plate assay, in the total material and within morphotypes.
|
Chicken caecal samples
|
Chicken retail meat samples
|
|
Morpho-type
|
No
|
%
|
Mean A595 ± SD
|
No
|
%
|
Mean A595 ± SD
|
p -value*
|
RDAR
|
57
|
67.1
|
1.677 ± 1.002
|
53
|
72.6
|
1.320 ± 0.895
|
0.04
|
BDAR
|
25
|
29.4
|
0.624 ± 0.757
|
17
|
23.3
|
0.591 ± 0.797
|
0.25
|
PDAR
|
3
|
3.5
|
0.100 ± 0.140
|
2
|
2.7
|
0.209 ± 0.064
|
0.80
|
SAW
|
0
|
0.0
|
|
1
|
1.4
|
0.083
|
-
|
ALL
|
85
|
100.0
|
1.312 ± 1.057
|
73
|
100.0
|
1.103 ± 0.924
|
0.16
|
SD = standard deviation
* Mean A595 of caecal samples vs retail meat samples
Single and mixed biofilms on glass slides
Six strains two of each of the morphotypes RDAR, BDAR and PDAR,elatively large differences in the total number of cfu in the biofilms when grown as single biofilms on glass slides. Furthermore, all except RDAR-1 had similar numbers of cfu in planktonic phase in the growth medium after incubation (Table 4). The number of cfu in the biofilms was not correlated to the number of planktonic cfu.
Table 4. Mean log of total cfu with standard deviation for each strain in the biofilm and in the planktonic fraction in the glass slide assay.
Strain no.
|
Source
|
Phylo-type
|
Strain name
|
Morpho-type
|
Biofilm log cfu
|
Planktonic log cfu
|
2014-01-2363
|
Caecal
|
A
|
RDAR-1
|
RDAR
|
5.82 ± 0.04
|
8.65 ± 1.34
|
2014-01-6040
|
Caecal
|
D
|
RDAR-2
|
RDAR
|
6.07 ± 0.16
|
9.59 ± 0.02
|
2014-01-5914-1
|
Meat
|
B2
|
PDAR-1
|
PDAR
|
6.31 ± 0.15
|
9.17 ± 0.34
|
2014-01-7342-1
|
Meat
|
B2
|
PDAR-2
|
PDAR
|
6.12 ± 0.0
|
9.34 ± 0.03
|
2014-01-7046-1
|
Caecal
|
A
|
BDAR-1
|
BDAR
|
4.44 ± 0.06
|
9.46 ± 0.12
|
2014-01-2069-1
|
Caecal
|
B1
|
BDAR-2
|
BDAR
|
5.07 ± 0.66
|
9.47 ± 0.12
|
Twelve different pairs were made by combinations of these strains to be used in studies on dual-species biofilms (Table S3). Each pair contained strains with different morphotypes. Both within the biofilm and in the planktonic phase, the mean total log10 cfu after incubation was higher when testing pairs of strains than when testing single strains, despite the total inoculum being the same in both (Table S4). The difference was statistically significant in the planktonic phase (p = 0.01), but not in the biofilm (p = 0.17). None of the strains displayed higher numbers of cfu within mixed than in the single biofilms. In fact, the mean log10 cfu for the PDAR strains, as well as for the RDAR-1 strain, were significantly lower in mixed biofilms with other morphotypes than in single strain biofilms (Table S5).
In mixed biofilms, BDAR strains constituted less than 10 % of the total cfu in all pairs, regardless of the ratio of the inoculum (Figures 1, 2). In these pairs, the ratio in the biofilm corresponded with the strains’ relative biofilm production as single strains. The ratios observed in pairs with one RDAR and one PDAR strain depended on which RDAR strain was included. The strain RDAR-2 dominated in both pairs with PDAR-strains (Figure 3), although these three strains all produced biofilms of the same magnitude when inoculated alone. In the biofilms where PDAR-strains were paired with the strain RDAR-1, the ratio seemed to reflect a combination of inoculation ratio and relative biofilm forming abilities as single strains (Figure 4).
In the planktonic phase, the cfu ratio of each pair of strains after incubation reflected the ratio inoculated (Figure 1-4). In addition, the BDAR strains outcompeted the PDAR strains, and the two RDAR strains outcompeted both the PDAR and the BDAR strains even though the strain RDAR-1 had the lowest growth rate when inoculated alone.
Single and mixed biofilms on CR agar plates
All six strains tested displayed typical morphotypes when grown as single biofilms on CR agar plates (Figure 5). The RDAR and PDAR strains, which both produced cellulose as part of the matrix, displayed highly structured biofilms with wrinkles and ridges. RDAR and BDAR strains, which both had curli fimbriae in the matrix, produced dark red/brown biofilms.
All inoculated pairs formed biofilms with a combination of structure and color indicating the presence of both cellulose and curli fimbriae during the first three days. After this time, the strains started expanding in a spatial fashion displaying their original morphotypes. For all pairs except those including RDAR-2, the ratio of the strains during spatial growth reflected the ratio that was displayed in the planktonic phase in the glass slide assay (Figure 6A). In contrast, the RDAR-2 strain dominated in all pairs, just as it did in biofilm on glass slides (Figure 6B).