The aim of this study was to investigate the effects of genomic DNA purified from Candida albicans and pneumonia-related pathogens, Pseudomonas aeruginosa and Staphylococcus aureus, on in vitro biofilm formation and morphological change of 3 Candida species (C. albicans, C. glabrata, and C. tropicalis).
Methods and Results
Biofilm formation was evaluated by the crystal violet assay and colony-forming unit counts. Morphological characteristics of biofilms were evaluated by scanning electron microscopy and fluorescence microscopy. Addition of DNA at a low concentration (<1·0 μg ml−1) significantly increased biofilm mass of all three Candida species. In contrast, the addition of DNA at a high concentration (10 μg ml−1) decreased the biofilm mass. Interestingly, the formation of hyphae in a dense network of yeast cells was observed in C. albicans biofilms exposed to a low concentration of DNA (<1·0 μg ml−1).
These findings demonstrated that extracellular DNA (eDNA) plays a crucial role in Candida biofilm formation and suggested that eDNA may induce the morphological transition from yeast to hyphal growth form during C. albicans biofilm development.
Significance and Impact of the Study
A novel therapy targeting eDNA may be applicable for Candida infection to decrease biofilm formation and hyphal formation.
Candida albicans, a predominant fungus in humans, can cause a variety of infectious diseases ranging from denture stomatitis in denture wearers (Radford et al. 1999; Ramage et al. 2004) to life-threatening invasive infections, particularly in immunocompromised and elderly populations (Nguyen et al. 1996; Bassetti et al. 2011). Among the factors contributing to the C. albicans pathogenicity, morphological transition between the yeast and hyphal growth forms and biofilm formation are considered as a major virulence factor (Felk et al. 2002; Nobile and Mitchell 2006). The hyphal transformation from yeast of C. albicans is influenced by environmental conditions (Odds 1985; Holmes et al. 1991).
The incidence of infections caused by Candida spp. is increasing worldwide, with a reported shift toward infections caused by Candida glabrata and C. tropicalis, which are commonly identified in samples from infected sites (Silva et al. 2012). C. tropicalis has recently emerged as a clinically important pathogen, more often associated with deep fungal infections (Zaugg et al. 2001). Following the increased use of immunosuppressive and broad-spectrum antibiotic treatments in modern medicine, the incidence of superficial and deep fungal infections associated with C. glabrata has also increased significantly (Hajjeh et al. 2004).
Candida spp. are frequently implicated in mixed bacterial–fungal infections; in particular, their presence in the respiratory tract of patients in intensive care units could facilitate bacterial growth and promote the development of ventilator-associated pneumonia caused by Pseudomonas aeruginosa and Staphylococcus aureus (Hamet et al. 2012). Candida albicans and Ps. aeruginosa were also co-isolated from the sputum of patients with cystic fibrosis (Bakare et al. 2003; Brand et al. 2008). Interestingly, extracellular DNA (eDNA) is also abundant in lung mucus of cystic fibrosis patients, reaching concentrations as high as 4 mg ml−1 (Palchevskiy and Finkel 2006).
The formation of bacterial biofilms has been extensively investigated (Whitchurch et al. 2002; Dell'Anno and Danovaro 2005; Hall-Stoodley et al. 2008); however, fungal biofilms have received less attention, and studies focusing on eDNA in fungal biofilms are few. In bacterial biofilms, the extracellular matrix (ECM) is mainly composed of highly hydrated proteins, polysaccharides and eDNA (Flemming et al. 2007). Recently, the presence of measurable amounts of eDNA in Candida biofilms has been discovered (Paramonova et al. 2009; Martins et al. 2010). It has been shown that eDNA is a crucial component of C. albicans biofilm ECM, contributing to the biofilm structural integrity and tolerance to antifungal drugs (Martins et al. 2010, 2011). In C. albicans, eDNA content in biofilm matrix increased with biofilm aging, adversely affecting biofilm strength (Paramonova et al. 2009). In this study, we investigated the effects of eDNA from C. albicans and pneumonia-related pathogens, Ps. aeruginosa and Staph. aureus, on biofilm formation and hyphal transformation of C. albicans, C. glabrata and C. tropicalis.
Materials and methods
Micro-organisms and growth conditions
Three strains of Candida, C. albicans clinical isolate CAD1 (Li et al. 2010), C. glabrata JCM3761 and C. tropicalis JCM1541 were used in this study. Each strain was inoculated in brain heart infusion (BHI) broth (Difco, Sparks, MD) and incubated at 37°C for 18 h under aerobic conditions (Hirota et al. 2005). After the culture reached a stationary phase, the Candida cells were suspended in yeast nitrogen base (YNB)/100 mmol l−1 glucose medium supplemented with 2·5 mmol l−1N-acetylglucosamine at a final concentration of 1·0 × 106 CFU ml−1 for biofilm experiments.
Staphylococcus aureus 209P and Ps. aeruginosa PAO1 were cultured in BHI and LB broth, respectively, at 37°C for 18 h under aerobic conditions.
Purification of genomic DNA, plasmid DNA and salmon sperm DNA
Candida albicans CAD1 genomic DNA was purified using lyticase (Sigma-Aldrich, St. Louis, MO) and NucleoSpin® Tissue kit (Macherey-Nagel, Duren, Germany) according to the manufacturers' instructions. Purity and concentration of the purified DNA were assessed by measuring the absorbance at 260 and 280 nm and by agarose gel electrophoresis. Genomic DNA from Ps. aeruginosa PAO1 and Staph. aureus 209P was purified as previously described (Nur et al. 2013). Plasmid DNA was isolated from overnight cultures of Escherichia coli XL1-Blue harbouring with pBR322, size of which is 4·361 kb, using GenElute Plasmid Miniprep kit (Sigma-Aldrich). The average size of commercially available sheared salmon sperm DNA solution (Life Technologies, Carlsbad, CA) was ≤2000 bp.
Biofilm formation and DNA treatment
Mucin from bovine submaxillary glands (M3895-Type I-S; Sigma-Aldrich) was dissolved in phosphate-buffered saline (PBS; pH 7·2) to a concentration of 0·5 mg ml−1. Then, type-I collagen-coated cell desks for 24-well plates (13·5 mm round, Celltight C-1 Celldesk LF; Sumitomo Bakelite Co., Tokyo, Japan) were immersed in mucin solution, and mucin adsorbance rate was quantified as described previously (Li et al. 2012).
Candida biofilms were grown on mucin-coated cell desks placed in the wells of flat-bottomed 24-well polystyrene cell culture plates as described previously, with a minor modification (Yoshijima et al. 2010). Briefly, 1·0 ml of Candida cell suspension in YNB/100 mmol l−1 glucose medium supplemented with 2·5 mmol l−1N-acetylglucosamine at a cell density was inoculated into the well containing cell desk. The plates were incubated for 90 min under aerobic conditions in an orbital shaker at 75 g to allow cell adhesion. After an initial adhesion phase, the medium was aspirated, and the cell desks were gently rinsed twice with 500 μl of PBS to remove any loosely adhered cells. Then, 1·0 ml of fresh YNB medium supplemented with purified DNA, plasmid pBR322 DNA or sheared salmon sperm DNA (0·01–10·0 μg ml−1) was added into the well, and the plates were incubated at 37°C for 24 h under aerobic conditions. After 24-h incubation, the medium was replaced with the fresh medium containing the purified DNA, and the plates were incubated for additional 24 h. The biofilms grown in YNB medium without DNA were used as a control. To determine the inhibitory effect of DNase I on biofilm formation, DNase I (Roche, Mannheim, Germany) was added to initially adhered cells or 24-h biofilms at a final concentration of 200 U ml−1. Biofilms grown in YNB medium without DNase I were used as a control.
Crystal violet assay
Biofilm mass was quantified using the crystal violet assay. Briefly, biofilms developed on the type-I collagen-coated cell desks were gently washed twice with 500 μl of PBS and stained with 300 μl of 0·1% aqueous crystal violet for 15 min. After staining, the cell desks were gently rinsed twice with 500 μl of PBS and transferred into new wells. The adsorbed dye was solubilized with 400 μl of 96% ethanol, and 50 μl of the solubilized solution was transferred into 96-well plate to measure the optical density at 540 nm using a microplate reader (model 680; Bio-Rad, Hercules CA). To analyze the biofilm stability, the formed biofilm mass before washing and the retained biofilm mass after washing were quantified by the crystal violet assay. The biofilm removability was calculated using the following equation: biofilm removability = (OD540 nm before washing − OD540 nm after washing)/OD540 nm before washing) × 100%.
Colony-forming unit (CFU) measurement
CFU counting method was used to quantify viable adherent Candida cells on each cell desk. Biofilms were gently rinsed with 500 μl of PBS and treated with PBS containing 0·25% trypsin to enzymatically detach the cells from the surface (Li et al. 2012). Following serial 10-fold dilution with PBS, 100 μl of suspension was plated on Sabouraud glucose agar, incubated at 37°C for 24 h under aerobic conditions, and the colonies were counted.
Scanning Electron Microscopy (SEM)
Biofilm structure was observed by SEM. Biofilms developed as described above were washed gently with distilled water and fixed in 2·5% glutaraldehyde for 1 h at 37°C. The wells were then rinsed three times with PBS, and the biofilms were dehydrated in increasing concentrations of ethanol up to 100% (Li et al. 2012). Dried biofilm samples were coated with Au and observed using the Miniscope TM-1000 (Hitachi High-Technologies Corp., Tokyo, Japan). SEM digital micrographs were taken and the total area of hyphae-type cells was determined from on-screen images using imagej (NIH, Bethesda, MD).
DNA labeling and fluorescence microscopy (FM)
Morphological characteristics of biofilms developed on type-I collagen-coated cell desks after 48-h incubation were evaluated by FM. Our recent study has demonstrated that 7-hydroxy-9H-(1,3-dichloro-9,9-dimethylacridin-2-one) (DDAO; Invitrogen, Carlsbad, CA) is useful for eDNA-specific staining, not intracellular DNA (Nur et al. 2013). Therefore, the biofilms were stained with 30 μl of 2 μmol l−1 DDAO for 30 min at room temperature in the darkness to stain eDNA, gently rinsed twice with 500 μl of PBS and stained with 10 μg ml−1 of Hoechst 33342 (DOJINDO, Kumamoto, Japan) for another 30 min in the darkness to stain intracellular DNA. After washing twice with 500 μl of PBS, each stained biofilm-containing cell desk was carefully placed on a glass slide with mounting medium and covered with a cover glass. The stained biofilms were observed using the Keyence BZ-9000 fluorescence microscope (Keyence, Osaka, Japan).
All data were statistically evaluated by one-way and two-way analyses of variance (anova) with Tukey's post hoc test for multiple comparisons using spss software (version 12.0; SPSS Japan Inc., Tokyo, Japan). The differences were considered significant when the probability value (P) was <0·01.
Effect of DNase I treatment on Candida biofilm formation
To elucidate possible effects of eDNA on biofilm formation, 200 U ml−1 of DNase I was added to initial adhered Candida cells and the cells were incubated for 48 h. As shown in Fig. 1, compared with untreated biofilms formed by all three Candida species, DNase I addition after the initial adherence phase significantly reduced biofilm mass at 48 h. The reduced OD540 levels by DNaseI treatment were 39·3 ± 0·8% for C. albicans, 30·3 ± 5·7% for C. glabrata and 31·4 ± 3·4% for C. tropicalis biofilms. These results suggest that eDNA plays an important role in Candida biofilm formation.
Effect of homologous eDNA addition on Candida biofilm formation and its stability
To further understand the effect of eDNA on Candida biofilm formation, initial attached Candida cells were incubated for 48 h to form a biofilm in the presence of homologous eDNA from Candida. The addition of homologous genomic eDNA significantly induced biofilm formation in a dose-dependent manner up to 0·1 μg ml−1 by crystal violet assay (Fig. 2a). The largest increase in biofilm mass was for C. albicans (74·5 ± 12·9% increase), followed by C. tropicalis (66·3 ± 1·0%) and C. glabrata (95·6 ± 9·0%). On the contrary, eDNA at a higher concentration of 10·0 μg ml−1 had a reducing effect (up to 43·7 ± 4·8%) on biofilm formation by all three Candida species (Fig. 2a).
To further determine the effects of Candida eDNA on its biofilm formation, 200 U ml−1 of DNase I was added to initial adhered Candida cells and incubated during biofilm formation for 48 h. As shown in Fig. 2b, compared with untreated biofilms formed by C. albicans, DNase I addition after the initial adherence phase significantly reduced biofilm mass at 48 h. This result also suggests that eDNA plays an important role during Candida biofilm formation.
To determine a possible effect of eDNA addition on the stability of the developed biofilms, we assessed the biofilm mass before washing and the retained biofilm mass after washing by crystal violet assay, and the biofilm removability was calculated (Fig. 2c). Interestingly, C. albicans biofilms formed in the presence of a higher eDNA concentration (10·0 μg ml−1) were broken off during washing and the biofilm removability was extremely high accordingly (51·1 ± 1·2%; Fig. 2c). However, the biofilm formed at a lower eDNA concentration (1·0 μg ml−1) was more resistant to fluid shear stress by washing (biofilm removability, 16·8 ± 1·4%). The same tendency in biofilm removability was observed for the biofilms formed by C. glabrata and C. tropicalis. These findings suggest that addition of eDNA up to 1·0 μg ml−1 can promote biofilm formation and strengthen biofilm structure, while higher eDNA concentrations negatively affect biofilm development and its stability.
Effect of homologous eDNA addition on Candida cell viability in biofilms
Next, we investigated a possible effect of homologous eDNA from Candida on cell viability in Candida biofilms. As shown in Table 1, the addition of eDNA up to 1·0 μg ml−1 did not significantly affect cell viability in biofilms formed by any Candida species. On the contrary, a significantly less number of colonies were observed in the biofilms formed in the presence of 10·0 μg ml−1 eDNA. These results suggest that eDNA at a higher concentration may negatively affect Candida cell viability in formed biofilm, whereas eDNA at a lower concentration has no effect on Candida cell viability in developed biofilm.
Table 1. Effect of homologous eDNA addition on Candida cell viability in 48-h biofilms
DNA (μg ml−1)
CFU ± SD (×108)
CFU, colony-forming unit.
Asterisks show significant differences vs the control biofilms (0 μg ml−1 DNA) (*P <0·01).
Effect of heterologous eDNA addition and small-sized DNA on Candida biofilm formation
We further observed that the addition of heterologous eDNA purified from pneumonia-related pathogens, gram-negative Ps. aeruginosa PAO1 and gram-positive Staph. aureus P209 affected Candida biofilm formation (Fig. 3a) and cell viability (Table 2) in a similar manner to homologous eDNA. These results suggest that eDNA, regardless of its origin, may influence the biofilm formation by Candida species in a dose-dependent manner.
Table 2. Effect of heterologous eDNA addition on Candida cell viability in 48-h biofilms
Pseudomonas aeruginosa PAO1
Staphylococcus aureus P209
DNA (μg ml−1)
CFU ± SD (×108)
CFU, colony-forming unit.
Asterisks show significant differences vs the control biofilms (0 μg ml−1 DNA) (*P <0·01).
Next, we determined the possibility that the size of eDNA affects biofilm formation by C. albicans, because the size of eDNA secreted from host and microbial species may differ from purified genomic DNA. Shorter DNA could also significantly induce Candida biofilm formation in a dose-dependent manner up to 0·1 μg ml−1, but shorter eDNA at a higher concentration of 10·0 μg ml−1 had a reducing effect on biofilm formation. These results suggest that the induction of biofilm formation by eDNA was not dependent on the length of the DNA fragments (Fig. 3b).
SEM and FM observations of Candida biofilm morphology
Finally, we examined the morphology of Candida biofilms formed on mucin-coated cell desks by SEM and FM (Fig. 4). The SEM images demonstrated that the 48-h biofilms of C. albicans CAD1 were fully mature consisting of a dense network of yeasts, pseudohyphae and hyphae. Interestingly, in the biofilms formed by C. albicans in the presence of 1·0 μg ml−1 eDNA, a dramatically increased hyphal growth was observed (Fig. 4a(a1–a3), d). However, biofilms of C. glabrata consisted exclusively of compact clusters of yeast cells, even in the presence of 1·0 μg ml−1 eDNA (Fig. 4b(b1–b3)). Candida tropicalis biofilms were formed mainly by yeast cells, although some long hyphal elements were observed (Fig. 4c(c1–c3)).
To further confirm distinct distributions of eDNA and intracellular DNA in formed Candida biofilm, we examined the 48-h Candida biofilms by FM after double staining with red fluorescent DDAO to identify eDNA and with blue fluorescent Hoechst 33342 to detect intracellular DNA. The biofilms formed in the presence of eDNA showed the morphological structure different from that of the DNA-free control biofilms (Fig. 4a–c). Fluorescent images of C. albicans biofilms formed in the presence of 1·0 μg ml−1 eDNA revealed a biofilm structure composed of a dense cell network embedded in ECM with high eDNA content in basal and superficial layers (Fig. 4a(a6,a7)). In a basal layer, granular clusters of eDNA were uniformly distributed among the biofilm cells, while more superficially localized eDNA demonstrated a definite cell- and hypha-shaped distribution, possibly due to impaired cell membranes of the stained cells. Hyphal elements within biofilms were intensely stained with the red fluorescent dye. The biofilms formed in the presence of 10·0 μg ml−1 of eDNA showed reduced blue fluorescence with blank areas due to diminished number of intact yeast cells and intensive red fluorescent spots from eDNA-aggregated ECM (Fig. 4a(a8,a9)). These observations confirm that eDNA at low concentrations promotes biofilm development and at high concentrations inhibits biofilm formation by C. albicans (Fig. 2). Therefore, the rinsing and washing during DDAO and Hoechst staining may cause this void space because of the decreased biofilm stability as shown in Fig. 2c, demonstrating Candida biofilms formed in the presence of a higher eDNA concentration (10·0 μg ml−1) were broken off during washing. Similar structures of biofilms formed by C. glabrata and C. tropicalis in the presence of eDNA were observed (Figs. 4b,c).
Moreover, we analyzed the average percentage of area occupied by hyphae-type cells, and this result showed that the biofilms formed by C. albicans in the presence of 1·0 μg ml−1 eDNA demonstrated significantly more hyphae-type cells' occupation compared to control biofilm without eDNA (Fig. 4d).
In this study, we employed a previously established method for biofilm formation on mucin-coated type-I collagen cell desks with minor modifications (Li et al. 2012). By using this biofilm model, we demonstrated that the addition (up to 1·0 μg ml−1) of eDNA from C. albicans and pneumonia-related pathogens, Ps. aeruginosa and Staph. aureus, promoted biofilm formation by C. albicans, C. glabrata and C. tropicalis, and the treatment with DNase I decreased biofilm mass (Figs 1-3), suggesting that eDNA as a component of the biofilm ECM is involved in the process of biofilm formation and subsequent maintenance. These findings are in accordance with previous result demonstrating that eDNA was an important component of C. albicans mature biofilms (Martins et al. 2010). Moreover, fluorescent microscopic images revealed that C. albicans biofilms formed in the presence of 1·0 μg ml−1 eDNA are composed of a dense cell network embedded in ECM with high eDNA content in basal and superficial layers (Fig. 4a(a6,a7)). Unlike the effect of eDNA at a low concentration, a high eDNA concentration (10·0 μg ml−1) significantly reduced both biofilm mass and its stability as well as cell viability (Fig. 2, and Tables 1 and 2), suggesting that high concentrations of eDNA may either inhibit the growth of Candida in biofilms and/or promote the detachment of viable cells from the biofilm surface leading to colonization in new environmental niches by detached cells. These findings correspond to previous data showing that DNA at concentrations <1·0 μg ml−1 induced biofilm formation by Streptococcus intermedius, while high DNA concentrations (>10 μg ml−1) negatively affected biofilm stability (Nur et al. 2013). In addition, shorter DNA could also significantly induce Candida biofilm formation in a dose-dependent manner up to 0·1 μg ml−1, suggesting that the induction of biofilm formation by eDNA was not dependent on the length of the DNA fragments (Fig. 3b). However, this result is in disagreement with the previous report showing ECM formation by Aspergillus fumigatus was dependent on the length of the DNA fragment (Shopova et al. 2013). The divergence between our result and previous study may reflect difference in fungal types.
Candida albicans is a dimorphic fungus that is capable of switching between the yeast and hyphal growth modes (Calderone and Fonzi 2001; Berman and Sudbery 2002). Serum, temperature, pH and Ca2+ are known as major external factors affecting this morphologic transformation (Odds 1985; Holmes et al. 1991). Hyphal forms are thought to play an important role in the pathogenesis of fungal infections due to their invasive growth (Felk et al. 2002). In our study, the extensive presence of hyphae revealed by DDAO staining was observed in the biofilms of C. albicans developed in the presence of a low eDNA concentration (1·0 μg ml−1) (Fig. 4a,d). Therefore, in Candida, eDNA in low concentrations may promote a morphologic shift from the yeast to hyphal growth mode.
In bacteria, it has been proposed that quorum sensing, i.e., regulation of gene expression in response to fluctuations in cell density is one of the mechanisms that contribute to release of DNA into environment (Spoering and Gilmore 2006). Furthermore, it has been considered that the mechanisms by which eDNA originates are two modes of autolysis: an altruistic suicide and a fratricide killing of different subpopulations of microbial cells (Montanaro et al. 2011). However, very little is known regarding the role of eDNA in formation of fungal biofilms. Recently, Paramonova et al. reported that C. albicans chk1/chk1, mutant strain, nonresponsive to the C. albicans quorum-sensing molecule farnesol because of the absence of histidine kinase Chk1p involved in the regulation of yeast to hyphae transition (Kruppa et al. 2004) increased eDNA levels but reduced biofilm strength, suggesting that eDNA may adversely influence biofilm integrity (Paramonova et al. 2009). Interestingly, the hyphae-to-yeast ratio affected the compression strength of C. albicans biofilms, and the biofilms with high hyphal content (>50%) were more resistant to vortexing and sonication (Paramonova et al. 2009). In our study, biofilms of C. albicans clinical isolate CAD1, developed in the presence of a low eDNA concentration (1·0 μg ml−1), showed extensive growth of hyphae and increased resistance to removal by fluid shear stress (Figs 2c and 4d).
Currently, C. glabrata represents the second most pervasive fungal pathogen after C. albicans and is capable of forming pseudohyphae and tubes (Fidel et al. 1999). In our study, biofilm mass formed by C. glabrata is fairly low, and the presence of true hyphae was not clearly observed in C. glabrata biofilms formed (Figs 2c and 3).
Candida tropicalis is considered an opportunistic yeast-like micro-organism that is common both in the environment and in normal human flora (Fromtling et al. 1987). It has been reported that some C. tropicalis isolates were more virulent than C. albicans isolates obtained from the same institution (Wingard et al. 1982). Our findings also showed that C. tropicalis biofilms were formed mainly with yeast-type cells although some long hyphae-type cells were also present (Fig. 4c(c1–c3)), suggesting that eDNA may affect the virulence of Candida species.
The presence of Candida spp. in the respiratory tract of patients in intensive care units could facilitate bacterial growth and promote the development of ventilator-associated pneumonia caused by Ps. aeruginosa and Staph. aureus (Hamet et al. 2012). Candida albicans and Ps. aeruginosa were also co-isolated from the sputum of patients with cystic fibrosis (Bakare et al. 2003; Brand et al. 2008), and eDNA is also abundant in lung mucus of cystic fibrosis patients, reaching concentrations as high as 4 mg ml−1 (Palchevskiy and Finkel 2006). Finally, we showed that heterologous DNA from pneumonia-related pathogens, Ps. aeruginosa and Staph. aureus, directly affected Candida biofilm development (Fig. 3a), suggesting that eDNA of all origins present at the infection sites may promote the formation of fungal biofilms and increase a severity of infectious disease, such as bacterial pneumonia.
To our knowledge, this is the first study showing that both homologous and heterologous eDNA directly affect Candida spp. biofilm development and its stability. The presence of eDNA in developing biofilms of C. albicans may influence biofilm structural integrity and cell growth mode; however, further studies are needed to elucidate the mechanism of eDNA effect on biofilm formation. Furthermore, targeting eDNA may be a novel strategy to battle infectious diseases, especially those associated with biofilm formation.
This work was supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (24592833 and 24592872).
Conflict of Interest
The authors declare that they have no conflicts of interest.