Extracellular DNA is essential for maintaining Bordetella biofilm integrity on abiotic surfaces and in the upper respiratory tract of mice

Abstract

Bacteria form complex and highly elaborate surface adherent communities known as biofilms which are held together by a self-produced extracellular matrix. We have previously shown that by adopting a biofilm mode of existence in vivo, the gram negative bacterial pathogens Bordetella bronchiseptica and Bordetella pertussis are able to efficiently colonize and persist in the mammalian respiratory tract. In general, the bacterial biofilm matrix includes polysaccharides, proteins and extracellular DNA (eDNA). In this report, we investigated the function of DNA in Bordetella biofilm development. We show that DNA is a significant component of Bordetella biofilm matrix. Addition of DNase I at the initiation of biofilm growth inhibited biofilm formation. Treatment of pre-established mature biofilms formed under both static and flow conditions with DNase I led to a disruption of the biofilm biomass. We next investigated whether eDNA played a role in biofilms formed in the mouse respiratory tract. DNase I treatment of nasal biofilms caused considerable dissolution of the biofilm biomass. In conclusion, these results suggest that eDNA is a crucial structural matrix component of both in vitro and in vivo formed Bordetella biofilms. This is the first evidence for the ability of DNase I to disrupt bacterial biofilms formed on host organs.

Figure 1. Bordetella biofilms contain eDNA.

Three day old biofilms of B. bronchiseptica strain RB50 (upper panels) and B. pertussis strain Bp536 (lower panels) were stained with DDAO. CLSM images of live GFP expressing cells (green) and DDAO stained eDNA (diffuse red) or dead cells (punctuate red) are shown. Yellow appearance indicates the presence of both live cells and eDNA. The images shown are representative of three independent experiments.

Figure 2. DNase I inhibits Bordetella biofilm formation.

The indicated strains were grown in 96 well microtitre plates for RB50 or 12 well plates for Bp536 for designated time points in SS medium supplemented with either DNase I resuspended in the reaction buffer or the reaction buffer alone. Wells were rinsed and stained with crystal violet followed by quantification of bound crystal violet as described in Materials and Methods. Each data point is the average of six wells, and error bars indicate the standard deviation. Representative data from one of at least three independent experiments are shown. Asterisks designate a value of P<0.05 (students t-test).

Figure 3. DNase I leads to the disruption of established Bordetella biofilms grown in microtitre plates.

Preformed 48h RB50 biofilms grown in 96 well plates were rinsed with PBS followed by incubation with PBS, PBS and reaction buffer, PBS and DNase I, PBS and heat inactivated (HI) DNase I, or PBS with reaction buffer and DNase I. Biofilm formation was then quantitated via crystal violet staining. Each point is an average of at least 6 wells, and error bars indicate the standard deviation. Asterisks designate a value of P<0.05 (students t-test).

Figure 4. DNase I leads to the disruption of established Bordetella biofilms grown on glass coverslips under static conditions.

Biofilms were grown on glass coverslips for 48h for RB50 (A) and 96 h for Bp536 (B). The coverslips were gently rinsed followed by treatment with DNase I for either 30min or 90min. The cells were tagged with GFP and thus are green. For each micrograph, the middle panel represents the x-y plane, and the adjacent top and side panels represent the x-z and y-z planes, respectively. The images of a biofilm not treated with DNase I and treated only with DNase I buffer are also depicted. CLSM was utilized to image the biofilms.

Figure 5. Susceptibility of flow cell biofilms to DNase I.

Representative z-reconstructions of RB50 biofilms grown under flow conditions for 6, 72, or 120h and imaged using CLSM for live GFP expressing cells (green) and eDNA stained with DDAO (red or yellow with co-localization). The image of untreated biofilms (left panels) were taken immediately prior to incubation with DNase I and the images of same biofilms treated with DNase I for 1.5h (left panels). Images shown here are representative of two independent experiments.

Figure 6. DNase I disrupts established biofilms of B. bronchiseptica and B. pertussis formed in the mouse respiratory tract.

CLSM images of biofilms harvested from mouse nose 15 and 19 days postinoculation with RB50 (top) or Bp536 (bottom), respectively. The harvested nasal septum was excised into two equal parts and incubated either with DNase I buffer (Mock, left panels) or with DNase I (right panels) before processing for staining as described in the Materials and Methods. Green staining depicts Bordetella biofilms formed on top of the host epithelium, which is stained red.

Figure 7. Scanning electron microscopy of mock treated (left) or DNase I treated Bordetella biofilms formed in the mouse nose.

Nasal septa were harvested from mice 15 days post-inoculation, excised into two equal parts, treated with either the DNase I buffer (Mock, left panels) or DNase I (right) followed by processing for SEM as described in the Materials and Methods.