Chitinases are negative regulators of Francisella novicida biofilms

Abstract

Biofilms, multicellular communities of bacteria, may be an environmental survival and transmission mechanism of Francisella tularensis. Chitinases of F. tularensis ssp. novicida (Fn) have been suggested to regulate biofilm formation on chitin surfaces. However, the underlying mechanisms of how chitinases may regulate biofilm formation are not fully determined. We hypothesized that Fn chitinase modulates bacterial surface properties resulting in the alteration of biofilm formation. We analyzed biofilm formation under diverse conditions using chitinase mutants and their counterpart parental strain. Substratum surface charges affected biofilm formation and initial attachments. Biophysical analysis of bacterial surfaces confirmed that the chi mutants had a net negative-charge. Lectin binding assays suggest that chitinase cleavage of its substrates could have exposed the concanavalin A-binding epitope. Fn biofilm was sensitive to chitinase, proteinase and DNase, suggesting that Fn biofilm contains exopolysaccharides, proteins and extracellular DNA. Exogenous chitinase increased the drug susceptibility of Fn biofilms to gentamicin while decreasing the amount of biofilm. In addition, chitinase modulated bacterial adhesion and invasion of A549 and J774A.1 cells as well as intracellular bacterial replication. Our results support a key role of the chitinase(s) in biofilm formation through modulation of the bacterial surface properties. Our findings position chitinase as a potential anti-biofilm enzyme in Francisella species.

Figure 1. Fn chitinase affects the biophysical properties of the biofilm.

Fn was grown to mid-log phase prior to the analyses. (A) The relative hydrophobicity of WT and chi mutants assayed by phenyl-sepharose column chromatography (HIC) and microbial adhesion to the nonpolar solvent hexadecane. *P<0.05, **P<0.01, and ***P<0.001 compared to WT (n = 6). (B) Autoaggregation of WT and chi mutants in PBS assayed at 24 and 48 h. *P<0.05, and **P<0.01 compared to WT (n = 6). (C) Size distribution for planktonic cultures of the strains in PBS measured by qNano analysis. (D) Particle translocation time (fwhm). The chi mutants had a larger fwhm duration than that observed for WT, indicating that the lower charge chi mutants took longer to traverse the pore. Mean presented in dots was calculated from every 100 data points.

Figure 2. Fn chitinase affects biofilm formation in different surface charged microplates.

(A) Biofilm formation based on CV staining (CV570) of cells adherent to negatively (TC), positively (Amine), neutral (PS) and positively/negatively (Primaria) charged 96-well plates, normalized by bacterial growth (OD600) expressed as CV570/OD600. (B) Attachment was assessed by CV staining 1 h post-inoculation of stationary-phase cultures (OD = 1.0). Initial attachment of Fn WT was very low to the TC and Primaria plates, but high to the amine and PS. *P<0.01 (n = 6) and NS (not significant) by unpaired Student's t-test.

Figure 3. COMSTAT2 analysis of WT and chi mutants.

Biofilms were grown in LabTek II glass chambers for 24-U confocal microscope. (A) 3D structures of biofilms were analyzed by CLSM z-stacks and z-stacks were rendered using Bitplane Imaris software. The images shown are representative of three independent experiments. (B) Mean thickness, (C) biomass, (D) surface to volume ratio, and (E) roughness coefficient of biofilms. *P<0.05 (n = 3) by unpaired Student's t-test.

Figure 4

(A) EPS contents of the cells and (B) culture supernatants of the strains. EPS contents were determined by phenol extraction followed by phenol-sulfuric acid method for carbohydrates as described in Materials and Methods. (C) Lectin binding assay to biofilms. FITC-Con A and FITC-WGA lectins were used for biofilm binding. Lectin binding capacity to biofilms was measured by a fluorescence plate reader and calculated relative fold to WT binding. Fluorescence microscopic images of biofilms of WT, chiA and chiB grown in TC plate are shown in the top panel. Biofilms in the TC plate were shown by CV staining (Fig. S1C). Scale bar, 100 μm.

Figure 5. Enzymatic activity of chitinase is required for regulation of Fn biofilm formation.

(A) Effect of exogenously added chitinase on biofilm formation in the negatively-charged TC plates. EC₅₀s of exogenous chitinase to WT, chiA and chiB mutants were determined to be 0.65, 0.18, and 0.21 μg/ml, respectively (n = 6). (B) Effect of exogenous chitinase on biofilm formation in the positively-charged amine plates. EC₅₀s of chitinase to WT, chiA and chiB mutants were determined to be 87.46, 0.17, and 0.15 μg/ml, respectively (n = 6). (C) Detachment of Fn biofilms after exposure to proteinase K, chitinase and DNase I (50 μg/ml) in the TC plates. Untreated control CV₅₇₀ values were 0.149±0.032, 0.588±0.012, and 0.585±0.017 for Fn WT, chiA and chiB mutants, respectively. *P<0.01 and **P<0.001 compared to control without enzyme treatment (n = 6).

Figure 6. Effect of chitinase inhibitors SAN and DEQ on antibacterial and antibiofilm activity.

(A, B) Susceptibility of Fn WT and chi mutants to SAN (A) and DEQ (B). Survival percentage of bacteria was calculated by OD₆₀₀ measurements after 24 h incubation with various concentrations of SAN and DEQ in TSBC. The EC₅₀s (μM) were determined by GraphPad software as indicated in the bottom table. (C) Effect of chitinase inhibitors SAN and DEQ on biofilm formation. Biofilm formation (CV570/OD600) was calculated by normalization with bacterial growth in each concentration of inhibitors. *P<0.05 compared to untreated (NT) control (n = 4).

Figure 7. Chitinase alters drug susceptibility of Fn biofilms.

(A, C) Effect of chitinase on drug susceptibility of biofilms pre-formed in the TC plates to (A) gentamicin (Gm) and (C) ciprofloxacin (Cipro). (B, D) Effect of chitinase on drug susceptibility of biofilms pre-formed in the amine plates to (B) gentamicin and (D) ciprofloxacin. (E) Susceptibility of chitinase-pretreated biofilms to gentamicin. Biofilms were formed on Amine plates in the presence of chitinase (0, 0.2 and 2 μg/ml) for 24 h then Gm (2 μg/ml) was added to the biofilms for 24 h. The remaining bacteria were calculated by the relative bacteria to no Gm-treated control in each concentration of chitinase. *P<0.05 compared to no Gm-treated control (n = 3).

Figure 8. Abrogation of Fn chi genes enhances ability to adhere to, to invade to and to replicate in host cells.

(A) Comparison of the adhesive properties of Fn WT and chiA mutants to A549 cells. *P<0.001 compared to WT (n = 6). (B) Bacterial invasion to A549 cells assayed by gentamicin protection method. *P<0.01 compared to WT (n = 6). (C) Effect of exogenous chitinase on bacterial adhesion. The same number of chitinase-treated bacteria as untreated bacteria were subjected to adhesion assays. *P<0.05 compared to untreated control of each strain (n = 3). (D) Bacterial adhesion assays using planktonic and biofilm cultures. Values are expressed as fold-increase adhesion relative to the planktonic counterparts. *P<0.05 and **P<0.01 compared to WT (n = 3). (E) Intracellular replication of the bacteria in host cells. A549 cells were infected with either WT or chi mutants at 100:1 MOI. Colony-forming units (CFUs) were determined after recovering intracellular bacteria from A549 cell lysates at 0 h or 18 h after gentamicin treatment to infected cells. (F) CFUs recovered from A549 cells lysed at 18 h post infection were compared with CFUs recovered at 0 h time point to calculate fold replication rate (change in CFU/hr). *P<0.05 compared to WT. (G) Bacterial invasion to J774A.1 cells assayed by gentamicin protection method. *P<0.01 compared to WT (n = 3). (H) Intracellular replication of the bacteria in J774A.1 cells. *P<0.05 and **P<0.01 compared to WT (n = 3).