Pleurotus ostreatus biofilm-forming ability and ultrastructure are significantly influenced by growth medium and support type



Alessandro D'Annibale, Department for Innovation in Biological, Agro-Food, and Forestry systems, University of Tuscia, V. S. Camillo de Lellis snc I-01100, Viterbo, Italy. E-mail:



To investigate the effect of support and growth medium (GM) on Pleurotus ostreatus biofilm production, specific metabolic activity (SMA) and ultrastructure.

Methods and Results

Biofilms were developed on membranes covering a broad range of surface properties and, due to the applicative implications of mixed biofilms, on standard bacterial GM in stationary and shaken culture. Hydrophilic (glass fibre, Duran glass and hydroxyapatite) and mild hydrophobic (polyurethane, stainless steel, polycarbonate, nylon) supports were more adequate for biofilm attachment than the hydrophobic Teflon. Among the GM, sucrose–asparagine (SA) was more conducive to biofilm production than Luria–Bertani and M9. GM was more influential than support type on biofilm ultrastructure, and a high compactness was evident in biofilms developed on SA. Biofilms on Duran glass were more efficient than planktonic cultures in olive-mill wastewater treatment.


The main effects of support and GM variables and their binary interactions on both biofilm production and SMA were all highly significant (P < 0·001): thus, the magnitude of the effect of each variable strongly depended on the level of the other one.

Significance and Impact of the Study

There is a lack of basic information regarding physiology and ultrastructure of P. ostreatus biofilms. To our knowledge, this is the first attempt to fill this gap, thus representing a basis for future studies.


In nature, micro-organisms are seldom found in planktonic form. Conversely, they tend to interact with solid matrices and other microbes so as to develop complex systems termed ‘biofilms’. The formation of biofilms represents one of the most widespread growth strategies of microbiota in natural ecosystems (Sutherland 2001; Stoodley et al. 2002). Although filamentous fungi are ideal biofilm-forming candidates due to their ability to produce extracellular polysaccharides (EPS) (Gutiérrez et al. 1995), lectins (Tsivileva et al. 2004; Kobayashi et al. 2005) and hydrophobins (Armenante et al. 2010), there is scant information regarding this topic because these micro-organisms cannot accurately fit within the restrictive biofilm definitions based on bacterial models. Thus, a set of structural criteria for filamentous fungal biofilm formation has been recently proposed by Harding et al. (2009), including surface-associated growth, encasement of hyphae within a self-produced extracellular polymeric matrix and, finally, occurrence of a complex aggregated growth involving the presence of either multilayers or hyphal bundles.

The white-rot basidiomycete Pleurotus ostreatus has attracted considerable interest due to its remarkable capacity to degrade a wide array of persistent organic pollutants (Leonardi et al. 2007) and to perform an effective biosorption of several heavy metals (Xiangliang et al. 2005). The biofilm-forming capability of this species, alone (Wu et al. 2005) or in association with bacteria (Jayasinghearachchi and Seneviratne 2006a,b), has been reported. These findings have been shown to have important implications in agriculture (Jayasinghearachchi and Seneviratne 2004, 2006b) and in wastewater treatment (Ragunathan and Swaminathan 2004; Wu et al. 2005). However, no basic information is currently available on structural and physiological properties of P. ostreatus-based biofilm systems.

The filamentous growth mode of white-rot basidiomycetes does not enable the use of high throughput approaches, such as those relying on either 96- or 192-well plates. Unlike bacteria and yeasts, in fact, the standardization of inocula added to such low volume systems is hampered by the objective difficulties of dispensing mycelial suspensions, unless spores are used. Thus, in this study, a novel approach based on the use of 12-well polystyrene plates able to host a variety of circular membranes of different materials was adopted to investigate the impact of support and growth medium on biofilm production, specific metabolic activity and ultrastructure under static and orbital-shaken conditions.

In the attempt of elucidating the role of the support, a variety of hydrophobic (i.e. Teflon), mild hydrophobic (i.e. stainless steel, polyurethane, polycarbonate, nylon) and hydrophilic (i.e. hydroxyapatite and glass-based supports) membranes were used. In fact, both amounts and structure of Candida biofilms have been reported to be significantly affected by the properties of the surfaces to which they adhere (Hawser and Douglas 1994; Chandra et al. 2001). Among them, substratum surface tension describing the ‘energetic state’ of a surface has been suggested to be a major property affecting attachment strength of organisms, and the supports used in this study covered a wide range of this property. Due to the aforementioned applicative implications of mixed biofilms involving P. ostreatus, this investigation was performed on standard liquid media generally employed for bacterial growth (Sambrook et al. 1989; Tremaroli et al. 2008). To the best of our knowledge, this is the first study reporting ultrastructural characteristics of monospecific P. ostreatus biofilms. Finally, olive-mill wastewater, the liquid waste from olive oil manufacturing, was taken as a model system to perform a comparative evaluation of degradation performances of planktonic and selected P. ostreatus biofilms.

Materials and methods

Materials and growth media

Multiwell polystyrene TC plates with 12 wells were from PBI International. 3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) was obtained from Sigma (Milan, Italy). GF/A (GFA, pore Ø, 1·6 μm; thickness, 0·26 mm), GF/B (GFB, pore Ø, 1·0 μm; thickness, 0·68 mm) and GF/D (GFD, pore Ø, 2·7 μm; thickness, 0·68 mm) glass fibre filters were from Whatman (Milan, Italy). Duran 0 (D0, pore Ø, 160–250 μm; thickness, 2·5 mm) and Duran 4 (D4, pore Ø, 10–16 μm; thickness, 2·0 mm) glass filters were from VWR (Milan, Italy). Nylon (NY, pore Ø, 0·2 μm; thickness 115 μm), Teflon (TF, pore Ø, 0·2 μm; thickness, 65 μm) and polycarbonate (PC, pore Ø, 0·2 μm; thickness 10 μm) membranes were purchased from Albet Hahnemuehle (Barcelona, Spain). Hydroxyapatite (HA) discs were from Biosurface Technologies Corp. (Bozeman, MT, USA). Stainless steel 316L wire net (SSN, pore Ø, 26 μm, thickness, 50 μm) and polyurethane foam (PU, pore Ø 350 μm; thickness, 2 mm) were gifts of Dr. Claudio Perani (University of Tuscia, Viterbo, Italy). The chemical compositions of the Luria–Bertani (LB) and sucrose–asparagine (SA) media are reported elsewhere (Tremaroli et al. 2008), while that of the M9 minimal medium in Sambrook et al. (1989). Olive-mill waster (OMW), withdrawn from a three-phase plant located in Viterbo (Italy), was centrifuged (11 000 g, 20 min), sterilized by autoclaving (121°C, 20 min) and stored at −20°C until used. The main chemical characteristics of OMW were as follows (g l−1): chemical oxygen demand (COD), 70; total phenols, 6·5; total sugars, 21·4; pH, 4·9.

Micro-organism and inoculum production

Pleurotus ostreatus (Jacquin: Fr.) Kummer, strain ATCC 58052, was stored a 4°C and periodically subcultured on malt extract agar (MEA). Mycelium fragments from 10-day-old malt extract agar slants were suspended in sterile deionized water (5 ml) using a sterile Potter homogenizer. Erlenmeyer flasks (500 ml) containing 95 ml of malt extract glucose medium (MEG, 10 g l−1 glucose and 5 g l−1 malt extract) were inoculated with these cell suspensions (5 ml per flask) and incubated on a rotary shaker (150 rpm, 28°C) for 120 h. The precultures were homogenized with an Ultra-Turrax (IKA Labortechnik, Staufen, Germany) (two subsequent steps of 20 s each, at c. 7000 rpm), centrifuged (6000 g, 10 min) and finally suspended with deionized water to yield a biomass concentration of c. 2·5 g l−1 mycelium dry weight.

Biofilm cultures were developed on standard liquid media for bacterial growth, namely SA, M9 and tenfold diluted Luria–Bertani (LBD), the respective C and N contents (g l−1) of which were 9·15 and 0·42, 0·8 and 0·35, and 0·65 and 0·21, respectively.

Biofilm production

Preweighed solid supports were sterilized in autoclave (121°C, 30 min) and aseptically added to each well of the above-mentioned multiwell plates. Replicate wells to which no support had been added were used as the control. Fungal suspensions (c. 0·4 ml corresponding to 1·0 ± 0·1 mg d.w.) were gently applied to either the surface of each disc or control wells and the plates incubated for 90 min under stationary conditions (adhesion period). Then, 1·8 ml of LBD, SA or M9 was added to each well and the plates incubated for 7 days at 30°C under either orbital (25 rpm) or static conditions in an incubator at 60% relative humidity. At the end of the incubation, after removal of the growth medium, phosphate-buffered saline (PBS, 2·0 ml) was added to each well and the plate orbitally shaken (70 rpm for 5 min). This procedure was repeated twice to remove both planktonic and loosely bound biomass (PLBB) and the washings filtrated under vacuum through preweighed GFD filters (Whatman, Maidstone, UK). Each support was then removed from its well and placed in an oven at 65°C up to constant weight. The amount of both PLBB and biofilm was separately determined by gravimetry and the total biomass obtained from their summation. In control wells, the biofilm was detached by an 83·1830 flexible cell scraper (Sarstedt, Newton, NC, USA), transferred onto preweighed GFD filter, dried at 65°C up to constant weight and then gravimetrically determined. The efficiency of the detachment procedure was higher than 90% as assessed by the method of Jin et al. (2003). The biofilm/total biomass ratio was calculated as an index of biofilm yield.

Metabolic activity of fungal biofilms

For each support, triplicate 7-day-old biofilm systems were added with 1·4 ml of MTT solution (5 mg ml−1 PBS), previously equilibrated at 37°C, and incubated at the same temperature for 5 h. At the end of the incubation, the formazan produced was solubilized several times with a dimethyl sulfoxide/0·2 mol l−1 glycine buffer, pH 8·6, mixture (SM, 6 : 1, v/v) in an ultrasonic 220 bath (Branson Ultrasonics, Dansbury, CT, USA) for 15 min. The pooled formazan extracts were read at 578 nm (ε = 16900 l mol−1 cm−1). Data were expressed as specific metabolic activity, referred to unit dry mass of biofilm. Although the tetrazolium salt 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT) is conventionally used in biofilm investigations, due to the high water-solubility of its formazan, in this study, a significant adsorption onto the mycelial biomass of its reduced product was observed, thus requiring a solubilization step. Consequently, MTT was selected due to its markedly lower cost.

Scanning Electron Microscopy (SEM) analysis of biofilms

Fungal biofilms were previously fixed overnight at 4°C with 2·0% paraformaldehyde and 2·0% glutaraldehyde in 0·15 mol l−1 cacodylate buffer, pH 7·4 (buffer A) (Karnovsky 1965). After thorough washing with buffer A (i.e. three changes, each at 4°C for 30 min), they were postfixed for 90 min at 4°C with 1·0% (w/v) osmium tetroxide in buffer A (Kellenberger and Ryter 1958). Specimens underwent three sequential washings (each at 4°C for 15 min) with the same buffer and dehydrated in an ascending series of ethanol. Samples were dried by the critical point method using CO2 in a Balzers Union CPD 020 (Vaduz, Liechtenstein), attached to aluminium stubs using a carbon tape and sputter-coated with gold in a Balzers MED 010 unit (Vaduz, Liechtenstein). The observation was made with a JSM 5200 electron microscope (JEOL, Tokyo Japan).

Olive-mill wastewater (OMW) treatment

An agitated 12-well microtitre plate system with a working volume of 3·5 ml was used to comparatively investigate degradation efficiency of P. ostreatus planktonic and biofilm cultures under conditions of limited oxygen supply. Either P. ostreatus 7-day-old planktonic cultures or selected coeval biofilms, obtained as described above, were rinsed twice with PBS for 15 min prior to OMW (2·8 ml) addition. Incubation was performed under orbital shaking (50 rpm) at 30°C for 8 days at 60% relative humidity to minimize evaporation losses. Control wells containing either heat-killed biofilms or planktonic biomass were incubated in parallel with OMW. Samples, withdrawn at selected time points, were centrifuged (11 000 g, 10 min) and analysed for their COD, total phenols and total sugars residual contents, as reported elsewhere (D'Annibale et al. 2006). All culture data were corrected for evaporation losses using the ratio between initial volume and the measured volume at selected time points. Per cent reductions in COD, total sugars and total phenols were calculated from difference between their removals in active cultures and those observed in respective controls (see above). Laccase and Mn-peroxidase activities were determined as previously described (D'Annibale et al. 2006).

Statistical analysis

Two-way analysis of variance (anova) of biofilm production and specific metabolic activity as a function of the growth medium and the support type was conducted, and post hoc multiple pairwise comparisons were performed by the Fisher least significant difference (LSD) test (P ≤ 0·05). Unit-variance-scaled and mean-centred biofilm, planktonic and loosely bound biomass, and total biomass amounts and biofilm/total biomass ratios were also subjected to principal components analysis (PCA) by the use of the Statistica 8.0 software (Statsoft, Tulsa, OK, USA) where hydrodynamic conditions were introduced as a categorical variable at two levels (i.e. static and shaken). The possible presence of either moderate or strong outliers in observations was checked by the squared prediction errors (SPE) of residuals and Hotelling (T2) of t-scores, respectively (MacGregor and Kourti 1995). At the variable level, variable power (VP), defined as the explained standard deviation, was calculated by equation (1):

display math(1)

where SVj is the residual standard deviation of the jth variable and math formula is its initial standard deviation, which is equal to unity for all variables after soft scaling.


Biofilm production

The endpoint for biofilm production was set at 7 days after the inoculation because P. ostreatus planktonic cultures reached the end of the exponential phase at that time in all the tested growth media. In particular, planktonic biomass production on SA and M9 did not significantly differ after 7-day incubation (2·44 ± 0·09 vs 2·35 ± 0·18 g l−1), while a tenfold dilution with deionized water of LB (LBD) yielded a similar growth (2·15 ± 0·22 g l−1).

Widely variable biofilm amounts, ranging from around 1·0 to 14·9 mg dry weight and corresponding to 5·3 to 131·8 mg cm−2 support, respectively, were achieved as a function of the factor combinations. Biofilm production data were subjected to two-way anova as a function of the growth medium and type of support, and multiple pairwise comparisons were performed (P ≤ 0·05). To better highlight comparisons within the same hydrodynamic conditions, static and shaken cultures were treated separately. Under both hydrodynamic conditions, the main effects of both growth medium and support variables were highly significant (P < 0·001) (Table 1). However, the same level of significance was found for their binary interaction (Table 1), thus meaning that the effect's size of each variable strongly depended on the level of the other one.

Table 1. Two-way anova of biofilm production in Pleurotus ostreatus P-3004 after 7-day incubation at 28°C in static and shaken (25 rpm) cultures as a function of growth medium and support
Sources of variationStatic culturesShaken cultures
  1. DF, degrees of freedom; SS, sum of squares; MS, mean squares; F, Fisher–Snedecor coefficient; P, significance level.

Growth medium2130·1065·0557·84<0·0012240·83120·4133·57<0·001
Growth medium × support22225·8110·269·13<0·00122389·5417·714·94<0·001
Residuals4449·481·12  42150·653·59  
Total791015·0212·85  771713·5222·25  

Irrespective of both growth medium and hydrodynamic condition, biofilm amounts on PU were higher than in the majority of the tested supports with the sole exception of shaken cultures on SA where the best biofilm production was observed on GFB and D0 (Fig. 1). Biofilm production on HA and SSN was significantly affected by hydrodynamic conditions. In particular, agitated conditions appeared to be unfavourable to biofilm formation on these supports (Fig. 1d–f). In the control wells, the biofilm formation only took place under static conditions (average levels of 4·5 mg dry weight) and was not affected by the growth medium (Fig. 1a–c); under this condition, biofilm formation was only observed in the flat bottom rather than on side walls of the wells. Regardless of both the growth medium and the hydrodynamic conditions, TF was the least performing support. The general framework of multiple pairwise comparisons, within each hydrodynamic condition, is shown in Fig. 1.

Figure 1.

(a–f) Biofilm production by Pleurotus ostreatus grown for 7 days at 28°C under static conditions on LBD, M9 and SA media (a, b and c, respectively) and under orbital shaking on the same media (d, e and f, respectively) in the absence and in the presence of various supports. The abbreviations of the supports are as follows: control (no added support); PU, polyurethane; SSN, stainless steel net; GFA, GFB and GFD, GF/A, GF/B and GF/D; NY, nylon; TF, Teflon; D0, Duran 0; D4, Duran 4; PC, polycarbonate; HA, hydroxyapatite. Data are the mean ± standard deviation of three replicates. Multiple pairwise comparisons were performed by the Fisher LSD test (P ≤ 0·05): same lowercase and uppercase letters above bars denote the absence of statistically significant differences between supports within the same medium and between media within the same support, respectively.

Two-way anova conducted on the overall main effect of the variable growth medium, irrespective of the support type, showed that the SA and LBD were the most and the least adequate media, respectively, in supporting biofilm production and this was also found for both static and shaken cultures. Thus, although all growth media had previously yielded statistically similar levels of planktonic biomass after 7-day incubation in both stationary and shaken cultures (data not shown), their overall effect on biofilm production, irrespective of the other tested variable levels, was found to statistically differ.

A multivariate approach was then conducted by PCA to obtain a general overview on data structure and on a comparative impact of hydrodynamic conditions. Figure 2 shows that around 76·8% of variability was explained by the first two principal components (42·4 and 34·4%, respectively), and no outliers were detected as indicated by Hotelling and SPE analysis of scores. Figure 2a shows that scores from shaken and static cultures clearly segregated from each other along the first component. In addition, the observations related to the absence of support and to TF, NY, PC and HA in shaken cultures were located in the diagonally opposite quadrant to that of the variables biofilm and biofilm/total biomass ratio, indicating that they were negatively correlated with them (Fig. 2b). Scores from PU and GFB in static cultures, instead, were in the upper left quadrant, which corresponded to that where the biofilm and the biofilm/total biomass variables were found. A significant effect of hydrodynamic conditions on SSN was observed, because its relative scores from shaken and stationary cultures clustered in the right upper and left upper quadrants corresponding to the positions of the planktonic biomass and biofilm variables, respectively. Thus, biomass adhesion on this support was disfavoured in the presence of agitation, while it acted as an effective support under static conditions.

Figure 2.

(a, b) Principal components analysis showing scores plot of different support/growth medium combinations under both static and shaken conditions (a) and loadings plot of the variables biofilm, planktonic and loosely bound biomass, total biomass and biofilm/total biomass ratio; hydrodynamic conditions were introduced as a categorical variable at two levels, namely stationary and shaken (b). Suffixes ST and SH after the names of the supports denote stationary and shaken cultures, respectively. Per cent variability explained by each principal component (PC) is shown between round brackets after each axis legend. The values of each variable power, calculated according to equation (1), are reported between round brackets close to the variable label in plot (b). (image_n) Control_ST; (image_n) Control_SH; (image_n) Polyurethane_ST; (image_n) Polyurethane_SH; (image_n) GFD_ST; (image_n) GFD_SH; (image_n) Stainless steel net _ST; (image_n) Stainless steel net_SH; (image_n) GFA_ST; (image_n) GFA_SH; (image_n) GFB_ST; (image_n) GFB_ST; (image_n) Nylon_ST; (image_n) Nylon_SH; (image_n) Teflon_ST; (image_n) Teflon_SH; (image_n) Duran 0_ST; (image_n) Duran 0_SH; (image_n) Duran 4_ST; (image_n) Duran 4_SH; (image_n) Polycarbonate_ST; (image_n) Polycarbonate_SH; (image_n) Hydroxyapatite_ST; (image_n ) Hydroxyapatite_SH.

Metabolic activity of Pleurotus ostreatus biofilms

Table 2 provides a synopsis of the specific metabolic activities of the biofilms under study, inferred from their abilities to reduce a tetrazolium salt and referred to unit mass of biofilm. Two-way anova applied to specific metabolic activity data showed that main effects of both growth medium and support variables and their binary interactions were highly significant (P < 0·001). Thus, as already observed for biofilm production, the magnitude of the effect of each variable depended on the level of the other one. Under static conditions, the most metabolically active biofilms were those developed on TF, PC and NY (Table 2). A significant metabolic activity (from 58 to 80 μmoles of MTT formazan h−1 g−1 biofilm) confirmed the above-reported P. ostreatus ability to develop a biofilm on the polystyrene flat bottom of control wells under static conditions (Table 2). Biofilms developed on PU, high production levels of which had been observed under static conditions, exhibited low metabolic activity.

Table 2. Specific metabolic activity of Pleurotus ostreatus biofilms developed in the absence and in the presence of several circular membranes with different surface tensions and pore sizes in static and shaken cultures at 28°C for 7 days
SupportSupport propertiesBiofilm specific metabolic activity (μmol MTT formazan h−1 g−1 biofilm d.w.)a
Surface tension (mN m−1)Pore size (μm)Static culturesShaken cultures
  1. n.s., not available; SSN, stainless steel net.

  2. a

    Data are the mean ± standard deviation of three experiments. Multiple pairwise comparisons were performed by the Fisher LSD test. (P ≤ 0·05): same lowercase and uppercase letters denote the absence of statistically significant differences between row and column means, respectively.

Nonen.a.n.a.57·7 ± 5·7B,b86·8 ± 5·0D,c80·0 ± 4·1C,c0·0 ± 0·0A,a0·0 ± 0·0A,a0·0 ± 0·0A,a
Teflon19·50·2101·2 ± 6·6C,b117·5 ± 13·7E,b156·8 ± 15·2E,c4·1 ± 1·5A,a16·8 ± 4·0A,B,a2·2 ± 0·7A,a
Polycarbonate33·50·2198·4 ± 9·0E,c68·6 ± 7·3C,b67·9 ± 5·8C,b23·1 ± 9·0B,a75·8 ± 7·0D,E,b92·4 ± 14D,b
Polyurethane foam39·335022·0 ± 0·1A,a20·0 ± 0·4A,B,a18·2 ± 0·3A,a51·1 ± 7·5C,D,b29·3 ± 0·8B,C,a46·0 ± 8·3C,b
SSN-316 L40·92629·8 ± 3·2A,a49·0 ± 2·7B,a,b49·0 ± 1·2B,a,b106·4 ± 13·8F,c51·0 ± 7·8C,b89·5 ± 5·1D,c
Hydroxyapatite57·00·1548·0 ± 11·6A,B,b14·9 ± 7·7A,a31·4 ± 7·5A,B,a,b4·7 ± 0·2A,a4·5 ± 0·2A,a14·8 ± 0·2A,B,a
Nylon46·50·2116·4 ± 22·0D,b,c141·4 ± 16·0F,c116·2 ± 2·3D,b,c74·0 ± 18·8D,E,a,b82·2 ± 15·42E,a,b49·1 ± 13·4C,a
GFA64·51·671·3 ± 12·6B,b45·0 ± 4·1B,a,b41·4 ± 0·8B,a62·3 ± 7·1D,b61·4 ± 6·7C,D,a,b50·3 ± 7·8C,a,b
GFB64·51·043·5 ± 2·7A,B,c20·8 ± 4·1A,B,a,b16·1 ± 1·9A,a40·6 ± 1·4B,C,c44·9 ± 3·4C,c25·3 ± 3·3B,b
GFD64·52·745·3 ± 4·0A,B,b34·1 ± 4·6B,a49·2 ± 5·3B,b52·9 ± 1·6C,D,b24·2 ± 3·5B,a28·2 ± 3·5B,a
Duran 064·5160–25035·2 ± 2·6A,a,b29·4 ± 4·2A,B,a28·4 ± 4·7A,a99·6 ± 8·7F,c46·6 ± 7·1C,b23·3 ± 4·4B,a
Duran 464·510–1662·8 ± 7·7B,b36·1 ± 4·1B,a20·6 ± 3·1A,a77·8 ± 2·4E,c56·3 ± 9·8C,D,b27·8 ± 3·1B,a

In shaken cultures, SSN delivered the most active biofilms along with PC and NY, while those on TF and HA were the least active ones. Thus, the performances of both TF and SSN were highly dependent on hydrodynamic conditions.

The effect of growth media on the specific metabolic activity markedly differed from those observed for biofilm production. In particular, the main effect of LBD, calculated by anova, was found to yield a significantly higher metabolic activity than that observed with the other growth media (P < 0·001) and this was found under both hydrodynamic conditions.

SEM analysis of Pleurotus ostreatus biofilms

SEM analysis was conducted on a number of biofilms developed on hydrophilic (i.e. GFB and D4), mild hydrophobic (SSN) and highly hydrophobic (Teflon) supports. The ultrastructural analysis on each selected support was also investigated as a function of the SA and LBD growth media that were shown to be most and least conducive to biofilm production.

All the specimens showed the presence of a highly intertwined network of hyphae, regardless of the support (Fig. 3). The degree of aggregation and compactness of the biofilm developed on the same support, however, was greatly affected by the growth medium. For instance, the massive presence of hyphal bundles embedded within a polysaccharide matrix on SSN was only observed when the biofilm was developed on SA (Fig. 3c). Conversely, on the same support on LBD, a less compact albeit highly intertwined network of hyphae with interspersed empty channels was evident (Fig. 3d). At a higher magnification, the surfaces of hyphae in the biofilm developed on SSN in the presence of SA appeared to be rough and lined by filaments of EPS (inset, Fig. 3c), while those pertaining to the SSN–LBD combinations were either smooth or covered by acicular crystals (inset, Fig. 3d). The same effect on biofilm compactness due to the growth medium was also evident for both GFB and D4 with SA leading to a higher degree of aggregation than LBD (Fig. 3e vs 3f, respectively, and Fig. 3a vs 3b, respectively). Conversely, the biofilm structure on TF as a function of the growth medium exhibited an opposite response to that observed for SSN leading to a more compact network on LBD than on SA (Fig. 3h vs 3g, respectively).

Figure 3.

(a–h) Scanning electron micrographs (magnification ×350) of Pleurotus ostreatus biofilms grown on either SA or LBD media and supported on Duran 4 (a and b, respectively), stainless steel net (c and d, respectively), GF/B (e and f, respectively) and Teflon (g and h, respectively). Insets within a white frame in each panel show details at ×2000 magnification. Twin oxalate crystals, EPS sheath or filaments and clamp connections, are indicated, respectively, by the following white symbols: image_n, image_n and ↑.

In both media, however, hyphae of the biofilm that developed on TF showed a rough surface lined with EPS and characterized by abundant presence of either monoclinic or bipyramidal concretions (Fig. 3g,h). The highest abundance and even distribution of these crystals along the hyphal length, however, was observed in the biofilms developed on GFB with SA as the growth medium. Figures 4a,b clearly show that the mycelial adhesion to the support was mediated by the presence of either EPS filaments or aggregates as in the cases of D4 and SSN. Although the critical point drying did not provide EPS in a hydrated state, it was possible to detect its presence either as a sheath lining hyphae (Fig. 3h), as observed, for instance, for the biofilms on TF, or as a continuous layer where hyphae were immersed, as in the case of GF-B (Fig. 4c).

Figure 4.

(a–c) Scanning electron micrographs of Pleurotus ostreatus biofilms grown on SA and supported on Duran 4 (a, magnification ×350), stainless steel net (b, magnification ×750) and GF-B (c, magnification ×350). Plots (a, b) show zones of fungal adhesion to the support. Plot (c) shows on the left and the right side a plectenchyma-like structure and the glass fibre filaments of GF-B, respectively.

Olive-mill wastewater (OMW) treatment

The OMW depollution efficiencies of three 7-day-old biofilm systems were compared with those of coeval P. ostreatus planktonic cultures. To this aim, biofilms developed on hydrophilic, mild hydrophobic and hydrophobic supports (i.e. Duran 0, SSN and Teflon, respectively) were selected and a wastewater characterized by rather high initial COD and total phenols contents (i.e. 70 and 6·6 g l−1, respectively) was used. To facilitate comparisons, the removed amounts of COD, total sugars and total phenols were either referred to treatment time as rates or normalized to unit biomass as yields. Table 3 shows that best performances were obtained after OMW treatment with biofilm on Duran 0. In particular, per cent COD reduction and its respective removal rate and yield (CRR and ORY, respectively) amounted to 37·9%, 138·3 mg l−1 h−1 and 2·53 mg mg−1 biomass, respectively. Conversely, worst results were obtained with biofilm on Teflon. With the only exception of the Teflon biofilm, no significant differences between planktonic and biofilm cultures were instead observed as far as the amounts of total phenols removed and their relative removal rates and yields were concerned. Both laccase and MnP activities were found at detectable albeit low levels. Best activity peaks of the former and the latter were found in biofilms developed on SSN (65·3 nkat l−1) and Duran 0 (89·9 nkat l−1), respectively (Table 3).

Table 3. COD, total sugars and total phenols removals with relative parameters of process performance and maximal laccase and Mn-peroxidase (MnP) activities after 8-day treatment at 30°C with 7-day-old planktonic Pleurotus ostreatus cultures or coeval biofilms of the same species developed on Duran 0, stainless steel net (SSN) or Teflon. Data are the mean of triplicate experiments and their coefficients of variability were always less than 4%. Row means followed by the same superscript letter were not significantly different (P < 0·05) as determined by the Fisher LSD test
ParameterPlanktonic culturesBiofilm developed on
Duran 0SSNTeflon
  1. CRR, COD removal rate; TSR, total sugar removal rate; PRR, phenol removal rate; ORY, organic load removal yield (mg COD removed mg−1 biomass); TSRY, total sugars removal yield (mg total sugars removed mg−1 biomass); PRY, total phenols removal yield (mg total phenols removed mg−1 biomass).

  2. *The time (days) required to attain the enzyme activity peak is indicated between round brackets.

COD removal (%)22·3b37·9d28·5c17·6a
Total sugars removal (%)27·0b29·2c29·4c18·9a
Total phenols removal (%)58·8b57·8b55·9b37·4a
Laccase activity (nkat l−1)*33·7b (4)41·7c (8)65·3d (8)0a
MnP activity (nkat l−1)*78·2b (6)89·9c (6)73·2b (6)12·3a (8)
CRR (mg l−1 h−1)81·2b138·3d104·0c64·3a
TSR (mg l−1 h−1)30·1b32·6c32·7c21·1a
PRR (mg l−1 h−1)19·9b19·6b18·9b12·6a


There is a lack of basic information concerning the impact of key variables, such as type of solid support and growth medium on P. ostreatus biofilm formation, metabolic activity and ultrastructure. The elucidation of the roles of these variables and their interplay might improve the use of P. ostreatus biofilm systems, for which potential applications in agriculture and in environmental restoration have been suggested. The former applications include the formation of associations with bacterial partners leading to improved N fixation (Jayasinghearachchi and Seneviratne 2004), endophytic colonization (Jayasinghearachchi and Seneviratne 2006b) and rock phosphate solubilization (Jayasinghearachchi and Seneviratne 2006a). In the latter ones, instead, both metal biosorption and removal of persistent organic compounds often involve the use of devices where the fungus mostly grows as a biofilm attached to a variety of inert supports (Ragunathan and Swaminathan 2004; Wu et al. 2005).

In the present study, although the surface tension of the selected substrates widely varied from 19 to 64·5 mN m−1, as in the cases of Teflon and glass fibre (Becker et al. 1997), respectively, both biofilm production and specific metabolic activity were not correlated with this property, in agreement with other studies (Becker 1996; Becker et al. 1997). The lack of a definite relationship between biofilm formation and surface tension might be due to the versatility of adaptive mechanisms involved in fungal adhesion to supports, relying on either hydrophobins (Wösten 2001; Armenante et al. 2010), lectins (Kobayashi et al. 2005) and/or EPS (Di Bonaventura et al. 2006). In addition, once the support is exposed to the suspending medium, it undergoes a rapid modification due to the early adsorption of both extracellular microbial macromolecules and attachment of colonizing cells, which overshadow surface tension effects (Goupil et al. 1980; Becker 1996). Thus, the differences in material properties, which are visible in the early incubation phases, might be levelled off by longer incubation times (Becker 1996). Fungal adhesion might also be affected by other support properties, such as surface roughness and intraparticle porosity (Guillemot et al. 2006; Tsang et al. 2007), as found by Asther et al. (1990) for polyurethane where higher values of these properties than other supports correlated with higher biofilm amounts. In the present study, polyurethane foam was found to be highly conducive to biofilm formation and this behaviour was not generally affected by either growth medium or hydrodynamic conditions.

It is noteworthy that SA, which was the most conducive medium to P. ostreatus biofilm production, had a significantly higher organic C content and C/N ratio than M9 and LBD. An important hallmark of microbial biofilms is that their structural stability and architectural shape largely depend on the amount of extracellular polymeric substances produced by constituent cells acting as an embedding matrix and promoting cell adherence on the surface of the support (Harding et al. 2009). Moreover, a complex medium with a C/N ratio very close to that of SA (i.e. 20·5 vs 21·8, respectively) yielded the best EPS production in P. ostreatus ‘florida’ cultures (Rosado et al. 2003). Likewise, a malt extract medium with a C/N ratio similar to that of SA was found to best support EPS production in liquid cultures of a closely related species, namely Pleurotus pulmonarius (El-Dein et al. 2004); interestingly, in the same study, asparagine, the sole N source in SA, was found to best support EPS release. This amino acid exerted a significantly higher effect than other N sources on the mycelial film formation in submerged Lentinula edodes F-249 cultures and this effect was due to an increase in the fungal lectin activity (Tsivileva et al. 2004). Lectins, a multivalent class of carbohydrate-binding proteins, can form cross-links between polysaccharides and glycoproteins, thus leading to their precipitation from solution and are also produced by P. ostreatus (Kobayashi et al. 2005).

In white-rot fungi, it has been shown that EPS might provide protection from dehydration, act as storage of either carbon sources or extracellular lignin-degrading enzymes and be involved in hyphal adhesion to the substrate (Gutiérrez et al. 1995). In P. ostreatus, the main EPS fraction is mainly composed of a 1→3 β-glucan with 1→6 branchings the frequency of which amounted to around 33% (Gutiérrez et al. 1995, 1996). In the present study, a relative high proportion of EPS in biofilm composition and subsequent diffusional limitations might explain why the specific metabolic activity in the majority of cases did not correlate with sessile biomass (Table 2 vs Fig. 1). Although PU provided a good anchorage and a tortuous pathway for hyphal penetration due to its high roughness and intraparticle porosity, it constrained, at the same time, the diffusion of both oxygen and nutrients, thus leading to low metabolically active biofilms (Table 2). For the opposite reason, a high specific metabolic activity was observed for biofilms developed on TF, PC and NY under static conditions (Table 2) although these supports had been shown to be scarcely conducive to biomass attachment (Fig. 1).

In the present study, SEM observations showed that the structural criteria suggested by Harding et al. (2009) were satisfied with the supports under study. However, biofilm systems exhibited differences that appeared to be driven by the interactions between culture medium and support surface. For instance, while the SA medium appeared to promote a higher degree of biofilm compactness than LBD on hydrophilic (i.e. GFB and D4) and mild hydrophobic (SSN) supports, an opposite trend was found with the hydrophobic support Teflon.

The presence of acicular and either monoclinic or bipyramidal particles along the hyphal length observed in the present work is in agreement with other ultrastructural studies on basidiomycetes (Dutton et al. 1993; do Rio et al. 2008); these concretions have been shown to be due to crystals of calcium oxalate, consisting of either its monohydrate or dihydrate form (i.e. whewellite and weddellite, respectively). Pleurotus ostreatus is a known producer of oxalic acid in both high and low N media reaching concentrations as high as 28 mmol l−1 in the former (Shimada et al. 1997). This low molecular mass metabolite exerts a variety of physiological functions such as defence from metals toxicity via chelation (Shimada et al. 1997) and ancillary role in lignin degradation (Khindaria et al. 1994; Hofrichter et al. 1999; Mäkelä et al. 2010).

To highlight possible differences between planktonic and biofilm P. ostreatus cultures in wastewater treatment efficiency, an OMW sample characterized by high COD and total phenols contents was selected. Several white-rot species, in fact, have been shown to be often ineffective on wastewaters with organic load and/or phenolic contents exceeding certain threshold levels (Sayadi and Ellouz 1993; Jaouani et al. 2003). In particular, several white-rot species failed to decolorize and remove phenols from OMW when COD was higher than 60 g l−1 (Sayadi and Ellouz 1993). As for P. ostreatus, a large strain-dependent variability can be inferred from the literature as far as its depolluting ability of olive effluents is concerned (Flouri et al. 1996; Aggelis et al. 2002; Fountoulakis et al. 2002). In our study, planktonic cultures of the P. ostreatus ATCC 58052 strain were able to grow and to partially reduce OMW organic load albeit with lower efficiency than those of coeval biofilms developed on Duran 0 and SSN; with these biofilm systems, although the per cent COD removals were not high in absolute terms, the respective removal rates were similar to those observed in another study where OMW treatment had been performed in either stirred-tank or bubble column reactors with Panus tigrinus (D'Annibale et al. 2006). This suggests that the use of the aforementioned P. ostreatus biofilms in devices able to ensure adequate oxygen supply might lead to significantly higher performances.

In conclusion, this study represents an original contribution aimed to systematically investigate some factors affecting biofilm formation and related ultrastructure and metabolic activity in a relevant species, such as P. ostreatus. It might provide a valuable basis for future studies dealing with this fungus in mixed biofilms, the applicative implications of which have been extensively reported. To the best of our knowledge, the ultrastructural characteristics of P. ostreatus biofilms have not yet been investigated. The use of 12-well plates hosting supports of different materials enabled a medium throughput approach, the use of which might be easily extended to screening programmes involving other fungal species and supports.


We thank the Ministero dell'Istruzione dell'Università e della Ricerca (MIUR) that supported this work within the project 2008P7K379.