Introduction

Over the past century, the potential of microalgae and cyanobacteria has been extensively explored in a wide range of applications. Current applications for these photosynthetic microorganisms range from human and animal nutrition to pharmaceuticals and cosmetics production (Singh et al. 2005; Spolaore et al. 2006; Del Campo et al. 2007; Priyadarshani and Rath 2012). For many years, microalgae and cyanobacteria have been applied in wastewater treatment processes since their use allows (i) low operational costs, (ii) no organic carbon requirements, (iii) the decrease of CO2 emissions associated to wastewater treatment plants, (iv) the obtainment of an oxygenated effluent, (v) heavy metals removal in a safer way, (vi) the recycling of nitrogen and phosphorus into microalgal/cyanobacterial biomass, and (vii) the reduction in sludge formation (Aslan and Kapdan 2006; Ruiz et al. 2011; Cai et al. 2013; Wang et al. 2014). However, the major drawback associated to wastewater treatment using these microorganisms is the separation of the produced biomass from the treated effluent (Ruiz et al. 2011; Abdel-Raouf et al. 2012). Current methods used for microalgal harvesting include chemical flocculation, bioflocculation, gravitational sedimentation, filtration, electrocoagulation-flocculation, flotation, centrifugation, and their combination (Milledge and Heaven 2013; Gonçalves et al. 2015). However, these methods are time-consuming or expensive (Barros et al. 2015). In this sense, immobilization systems emerged as a good alternative, as they avoid further steps of biomass recovery from the effluent.

Microalgal/cyanobacterial biofilms have gained particular attention in wastewater treatment systems since they can promote a rapid and efficient harvesting of biomass, leading to lower downstream processing costs (Abdel-Raouf et al. 2012). There are many reactors using microalgal biofilms for wastewater treatment. The Algal Turf Scrubber is one of the most used microalgal biofilm reactors for wastewater treatment. This system was invented and patented by Adey (1982) and has been widely used ever since by others (Jensen 1996; Schumacher et al. 2003; Stewart and Zivojnovich 2003) with some modifications. Several studies have shown satisfactory results regarding nutrient removal efficiencies (36–93%) and biomass productivities (5 to 20 g m−2 day−1) using this system (Mulbry and Wilkie 2001; Wilkie and Mulbry 2002; Mulbry et al. 2005). Muñoz et al. (2009) compared a flat plate photobioreactor and a tubular photobioreactor, demonstrating that higher biochemical oxygen demand (BOD) removal rates can be obtained with microalgal biofilms (27 mg and 32 mg BOD L−1 h−1, respectively). More recently, Johnson and Wen (2010) developed a polystyrene rocker system using different materials for microalgal and cyanobacterial attachment. Although the aim of these authors was to enhance biofuel production, this mechanism allowed nutrient removal efficiencies ranging between 70 and 100%. Boelee et al. (2011) designed and operated a rotating algal biofilm reactor at laboratory and pilot scales in order to maximize microalgal production in wastewaters resulting from the secondary treatment step of wastewater. With this system, high nutrient removal rates were reached with values of 14.1 and 2.1 g m−2 day−1 for total nitrogen and phosphorus, respectively.

Biofilms are an assemblage of cells attached to a substrate and embedded in a self-produced matrix of extracellular polymeric substances (EPS) (Donlan 2002; Christenson 2011; Christenson and Sims 2011). An ideal substrate for biofilm formation should present the following characteristics: non-toxicity, stability, and ability for biomass retention (Mallick 2002). Microbial adhesion to a substrate constitutes one of the main steps of biofilm development and strongly depends on several factors that can be divided into: (i) biological factors, (ii) physicochemical properties, and (iii) environmental conditions. Biological factors influencing microbial adhesion and biofilm formation include EPS production (Zhang and Fang 2001) and the presence of external structures, such as fimbriae, prosthecae, pili, or flagella, able to promote the contact between the microorganisms and the adhesion substrate. Physicochemical properties of the adhesion surfaces also influence the microbial attachment process. Finally, environmental conditions affecting photosynthetic biofilms formation include light (Hill et al. 1995), pH (Liehr et al. 1990), hydrodynamic conditions (Kokare et al. 2009), nutrient quality and quantity (Sekar et al. 2002; Kesaano and Sims 2014), temperature (Babu 2011), and carbon source (Yang and Gao 2003). Several studies have focused on the effects of surface physicochemical properties on microbial adhesion and biofilm formation, reporting that in general, microbial adhesion tends to occur preferably on hydrophobic surfaces (Holland et al. 2004; Sekar et al. 2004; Stanley and Callow 2007; Li et al. 2010; Ozkan 2012). However, other factors, such as culture medium and microbial species, can play an important role on microalgal/cyanobacterial attachment to a surface (Becker 1996; Johnson and Wen 2010; Irving and Allen 2011).

Although there are many studies focusing on surface physicochemical properties of several microorganisms and surface materials, studies comparing thermodynamic approaches and in vitro biofilm formation experiments are not very common or even absent for microalgae and cyanobacteria (Sekar et al. 2004; Irving and Allen 2011; Ozkan and Berberoglu 2011; Sirmerova et al. 2013; Gonçalves et al. 2015;). Accordingly, the aim of this study was to determine the surface physicochemical properties of Chlorella vulgaris, Pseudokirchneriella subcapitata, Synechocystis salina, and Microcystis aeruginosa, as well as those of selected surface materials (copper—Cu, glass—G, poly(methyl methacrylate)—PMMA, polystyrene—PS, polyvinyl chloride—PVC, and AISI316 stainless steel—SS) and evaluate their influence on biofilm formation. The influence of different medium compositions was also assessed through biofilm formation in vitro experiments. These microorganisms were selected because they have been extensively used in several applications, namely in wastewater treatment processes (Moreira-Santos et al. 2004; Memon et al. 2014; Gonçalves et al. 2015. The surface materials were selected because: (i) Cu is a control surface due to its cellular toxicity (Stauber and Florence 1987; Trevors and Cotter 1990; Levy et al. 2007; Grass et al. 2011); (ii) Cu, PVC, and SS are widely used in industrial applications (Carvalho et al. 2006, Kim et al. 2013); (iii) G, PMMA, and PS are transparent surfaces, allowing light penetration, which can be relevant for the growth of photosynthetic organisms (Booth 1971; Knauss and Schapery 1999); and (iv) PMMA and PVC are commonly used as biofilm adhesion support in wastewater treatment processes (Wuertz et al. 2003; Muñoz et al. 2009; Kesaano and Sims 2014).

Materials and methods

Microorganisms and culture conditions

The microalgae Chlorella vulgaris CCAP 211/11B and Pseudokirchneriella subcapitata CCAP 278/4 were obtained from Culture Collection of Algae and Protozoa (UK), while the cyanobacteria Synechocystis salina LEGE 06079 and Microcystis aeruginosa LEGE 91344 were obtained from the Laboratory of Ecotoxicology, Genomic and Evolution – CIIMAR (Centre of Marine and Environmental Research of the University of Porto, Portugal). Stock solutions of these microorganisms were prepared in OECD (Organization for Economic Co-operation and Development) test medium (OECD 2011) with the following composition (per liter), 250 mg NaNO3, 12 mg MgCl2·6H2O, 18 mg CaCl2·2H2O, 15 mg MgSO4·7H2O, 45 mg KH2PO4, 0.08 mg FeCl3·6H2O, 0.1 mg Na2EDTA·2H2O, 0.185 mg H3BO3, 0.415 mg MnCl2·4H2O, 3 μg ZnCl2, 1.5 μg CoCl2·6H2O, 0.01 μg CuCl2·2H2O, 7 μg Na2MoO4·2H2O, and 50 mg NaHCO3. Growth medium was sterilized by autoclaving at 121 °C for 15 min. The cells were incubated in 500-mL flasks at room temperature (24.0 ± 1.0 °C), under continuous fluorescent light with an irradiance of approximately 120 μmol photons m−2 s−1 at the surface of the flasks (Gonçalves et al. 2015). Agitation was obtained by bubbling atmospheric air (filtered through 0.22-μm cellulose acetate membranes, Orange Scientific, Belgium) at the bottom of the flasks.

Surface contact angles

Microalgal and cyanobacterial suspensions in the exponential growth phase were harvested, washed twice, and resuspended in saline solution (0.85%, w/v, NaCl) to obtain a final concentration of about 5.0 × 106 cells mL−1. Microbial lawns were prepared by filtering the previously washed suspensions using 0.45-μm nitrocellulose membrane filters (Advantec MFS, Inc., Japan) until complete clogging of the membranes. The surface materials used in this experiment were Cu, G, PMMA, PS (bacterial grade), PVC, and SS. The surfaces were washed with commercial detergent and sterile water, and sterilized under UV light for 30 min. Contact angle measurements of the above-described lawns and the surface materials were performed using the sessile drop method, as described by Busscher et al. (1984). The measurements were carried out at room temperature using water, formamide, and α-bromonaphthalene (Sigma-Aldrich, Portugal) as reference liquids. Determination of contact angles was performed automatically using an OCA 15 Plus (Dataphysics, Germany) video-based optical contact angle measuring instrument that allows image acquisition and data analysis. Contact angle measurements (at least 20 determinations for each liquid and for each suspension/material) were performed in three independent experiments.

Surface physicochemical parameters and hydrophobicity

Surface hydrophobicity of the selected microorganisms and surface materials were determined using the approach of Van Oss (1995), which allows the assessment of the absolute degree of hydrophobicity of any surface in comparison with their interaction with water. These determinations were performed as described by Gonçalves et al. (2015).

Free energy of adhesion determinations

The free energy of adhesion between the selected microorganisms and surface materials was determined according to Simões et al. (2008). When studying the interaction between surfaces i (microalgal/cyanobacterial cells) and I (surface materials) that are immersed or dissolved in water, the total interaction energy (∆Gadhesion, in mJ m−2) can be assessed through the following expression:

$$ {\Delta \mathrm{G}}_{\mathrm{adhesion}}={\upgamma}_{\mathrm{i}\mathrm{I}}^{\mathrm{LW}}-{\upgamma}_{\mathrm{i}\mathrm{w}}^{\mathrm{LW}}-{\upgamma}_{\mathrm{I}\mathrm{w}}^{\mathrm{LW}}+2\left[\sqrt{\upgamma_{\mathrm{w}}^{+}}\left(\sqrt{\upgamma_{\mathrm{i}}^{-}}+\sqrt{\upgamma_{\mathrm{I}}^{-}}-\sqrt{\upgamma_{\mathrm{w}}^{-}}\right)+\sqrt{\upgamma_{\mathrm{w}}^{-}}\right]\times \left(\sqrt{\upgamma_{\mathrm{i}}^{+}}+\sqrt{\upgamma_{\mathrm{I}}^{+}}-\sqrt{\upgamma_{\mathrm{w}}^{+}}\right)-\sqrt{\upgamma_{\mathrm{i}}^{+}{\upgamma}_{\mathrm{I}}^{-}}-\sqrt{\upgamma_{\mathrm{i}}^{-}{\upgamma}_{\mathrm{I}}^{+}} $$
(1)

Thermodynamically, adhesion is expected to occur if ∆Gadhesion < 0 , while if ∆Gadhesion > 0, microbial adhesion is not favored.

Biofilm formation in vitro assays

In vitro biofilm formation assays were performed based on the method proposed by Meireles et al. (2015). In these experiments pieces of Cu, G, PMMA, PS, PVC, and SS were used as biofilm substrates. The dimensions of the pieces were 1.0 × 0.9 cm, with a thickness ranging from 0.1 to 0.2 cm. The pieces were washed with commercial detergent and sterile water, and sterilized under UV light for 30 min. Additionally, biofilm formation experiments were performed using two different media: OECD test medium (OECD 2011) and a synthetic medium simulating a domestic effluent (synthetic effluent, SE), both protein-free. The OECD test medium is an inorganic medium presenting some macronutrients, such as inorganic nitrate and phosphate, and a wide variety of inorganic salts, the trace elements required for microalgal/cyanobacterial growth. Composition of this culture medium is the same as the one described by Gonçalves et al. (2015). On the other hand, the SE used in this study is a modified version of the one proposed by Gebara (1999), having the following composition (per liter), 300 mg C6H12O6, 62 mg NaNO3, 10 mg MgSO4·7H2O, 11 mg KH2PO4, 1 mg MnSO4·H2O, 0.46 mg CaCl2, and 0.05 mg FeCl3·6H2O. This culture medium also contains inorganic sources of nitrogen and phosphorus. However, it contains an organic carbon source (glucose). The initial pH of the culture media was 7.1 and 6.7 for the OECD and SE test media, respectively.

After cleaning and sterilization, the coupons were inserted in 12-well microtiter plates (Orange Scientific, Braine-l’Alleud, Belgium). Microalgal/cyanobacterial suspensions of the selected microorganisms were centrifuged at 16800×g for 15 min (5810 R Eppendorf, Germany). After centrifugation, the supernatant was discarded and the suspensions were diluted in each of the culture media evaluated in this study to obtain a final concentration of 1.0 × 106 cells mL−1. Then, 3 mL of each cell suspension was added to the respective well with the coupon already inserted. The plates were placed in a KS 130 orbital shaker (IKA Werke, Germany) with constant agitation of 160 rpm. Biofilm formation was allowed to occur for 24, 72, and 168 h at room temperature and constant light intensity (approximately 120 μmol photonsm−2 s−1, as the one used for stock solutions preparation). After these periods, the coupons were placed in 15-mL tubes containing saline solution and microalgal/cyanobacterial cells were removed from the coupons by vigorously vortexing for 1 min. These suspensions were further used for quantification of biofilm cells. Additionally, the pieces containing the adhered cells were analyzed by scanning electron microscopy (SEM) to evaluate the organization of cells at the surface of the selected materials. Three independent experiments were performed for each coupon in each condition.

Quantification of biofilm cells

Biofilm cells were quantified through colony forming units (CFUs) determinations. To determine the number of CFUs, the necessary dilutions in saline solution were prepared and plated on Bold Basal medium (Hoff and Bold 1963) supplemented with agar (1.5%, w/v), using the motion drop method (Reed and Reed 1948). These assays were carried out in duplicate. The cells on the studied substrate pieces were expressed in terms of CFU per square centimeter, according to Eq. 2:

$$ \mathrm{CFU}\ {\mathrm{cm}}^{-2}=\frac{\frac{N}{V_{\mathrm{P}}}\bullet {V}_{\mathrm{S}}\bullet D}{A} $$
(2)

where N is the number of CFUs, VP is the volume plated (0.01 mL), VS is the volume of saline solution in which biofilms were resuspended, D is the dilution factor, and A is the total area of each substrate (cm2).

SEM observations

For SEM observations, the samples were fixed, successively dehydrated, and dried, according to the method described by Gomes et al. (2013).

Statistical analysis

The statistical analysis was performed using the statistical software SPSS 22.0 (SPSS Inc., USA) at a significance level of 0.05. The surface physicochemical parameters were analyzed using paired-samples t test. Furthermore, the influence of strain, culture media, and material, as well as a combination of these factors, in the different parameters studied were evaluated through three-way-ANOVA.

Results and discussion

Surface physicochemical properties of the selected microorganisms

Contact angles measurements allowed the determination of the surface tension parameters of C. vulgaris, P. subcapitata, S. salina, and M. aeruginosa. These results are presented in Table 1 in a decreasing order of free energy of hydrophobic interaction (from the most hydrophilic to the most hydrophobic). It is possible to observe that all microbial surfaces were hydrophilic (\( {\Delta G}_{sws}^{TOT}>0 \) mJ m−2). In a previous study of Gonçalves et al. (2015), aiming to evaluate the influence of surface physicochemical properties of microalgae and cyanobacteria on their sedimentation kinetics, similar \( {\Delta G}_{sws}^{TOT} \) values were obtained for C. vulgaris, P. subcapitata, and M. aeruginosa (52.5, 13.8, and 43.4 mJ m−2, respectively) using the OECD test medium for microbial growth. Ozkan and Berberoglu (2013c) studied the surface physicochemical properties of Synechocystis sp. and verified a hydrophilic character (\( {\Delta G}_{sws}^{TOT} \) of 10.9 mJ m−2) for this cyanobacterium, which corroborates the data obtained in this study for S. salina.

Table 1 Contact angles and surface physicochemical properties of the studied microorganisms and surface materials
Full size table

The Lifshitz van der Waals component, \( {\gamma}_s^{LW} \), was similar (p > 0.05) for all studied microorganisms, with values ranging from 35.8 ± 1.0 to 38.1 ± 1.8 mJ m−2. Previous studies (Ozkan and Berberoglu 2013c; Gonçalves et al. 2015) provided \( {\gamma}_s^{LW} \) values of 35.0 and 37.8 mJ m−2 for different microalgae and cyanobacteria. LW forces, usually attractive, result from instantaneous asymmetrical distribution of electrons in molecules (the higher the value of the LW component, the more apolar is the surface and, therefore, the lower would be its affinity for polar liquids) (Ozkan and Berberoglu 2013b). Accordingly, similar results obtained for this parameter indicate that cell wall composition of the studied microorganisms may be similar (Latala et al. 2009; Wurdack 1923). In the study performed by Ozkan and Berberoglu (2013c), differences in the LW component observed for green algae and diatoms were attributed to differences in their cell wall chemical composition. Electron donor and acceptor parameters give information about the properties of molecules present in the surfaces studied: higher \( {\gamma}_s^{+} \) indicate the presence of positively charged molecules and higher \( {\gamma}_s^{-} \) indicate the presence of negatively charged ones (Janczuk et al. 1993). In this study, all microorganisms presented electron donor properties, with the predominance of the electron donor component, \( {\gamma}_s^{-} \). This polar character may be due to the presence of excessive molecules of oxygen or polar groups on microbial surfaces (Van Oss 1995, 2006; Gonçalves et al. 2015). Additionally, the values determined for the studied microorganisms were not statistically different (p > 0.05), with the exception of M. aeruginosa. Regarding the acid-base or polar component, the values determined for the studied species were not statistically different (p > 0.05). In the study of Ozkan and Berberoglu (2013c), similar \( {\gamma}_s^{AB} \) values were obtained for Nannochloris sp., Nannochloris oculata, Chlorella vulgaris, Scenedesmus dimorphus, Afrocarpus falcatus, Botryococcus sudeticus, and Botryococcus braunii, ranging from 0.0 to 5.1 mJ m−2. Acid-base forces result from electron transfer interactions between polar components of the involved surfaces, which can be attractive (hydrophobic attraction) or repulsive (hydrophilic repulsion) (Gonçalves et al. 2015). In this study, the \( {\gamma}_s^{AB} \) component represents a measure of hydrophilic repulsion, since all microorganisms presented a hydrophilic character\( \left({\Delta G}_{sws}^{TOT}>0\right) \).

Surface physicochemical properties of the selected materials

The surface tension parameters were also determined for the selected surface materials. As for microbial surfaces, these results are presented in Table 1 in the decreasing order of free energy of hydrophobic interaction. The \( {\Delta G}_{sws}^{TOT} \) values obtained were statistically different (p < 0.05) between the studied materials, with the exception of the SS-PMMA, PS-G, PS-PVC, and PVC-G combinations. As it can be observed from Table 1, contact angles determined using the polar liquids on the studied surfaces (with the exception of Cu and PS) present values below 90°, meaning that surface-liquid interactions dominate the system. Only SS presented a hydrophilic character (\( {\Delta G}_{sws}^{TOT}>0 \) mJ m−2). In fact, surfaces presenting higher free energy of hydrophobic interaction are more hydrophilic, which is the case of SS (Visser 1995; Kokare et al. 2009). In contrast, Cu, G, PMMA, PS, and PVC are hydrophobic surfaces, due to the negative value determined for \( {\Delta G}_{sws}^{TOT} \). Similar \( {\Delta G}_{sws}^{TOT} \) values have already been reported for Cu (− 79.6 mJ m−2) and G (− 13.8 and − 14.8 mJ m−2) (Teixeira et al. 2005; Simões et al. 2007). However, it is important to notice that it is likely that Cu and G were probably contaminated with a hydrophobic material (e.g., hydrocarbon from the air) before the experiments. Some authors reported similar \( {\Delta G}_{sws}^{TOT} \) values for PMMA, with values of about − 18.9 (Teixeira et al. 2005) and − 16.8 mJ m−2 (Teixeira and Oliveira 1999; Oliveira et al. 2001). In the study performed by Lopes et al. (2005) PMMA was also determined as hydrophobic, but the absolute value of \( {\Delta G}_{sws}^{TOT} \) obtained was higher (36.3 mJ m−2) than the one determined in this study (19.1 ± 6.8 mJ m−2). The free energy of hydrophobic interaction determined for PVC (− 22.6 mJ m−2) is in accordance with the results of Teixeira and Oliveira (1999) and Oliveira et al. (2001), who reported \( {\Delta G}_{sws}^{TOT} \) values of about 22.0 mJ m−2. Nevertheless, it was found that \( {\Delta G}_{sws}^{TOT} \) values determined for PS were lower than those reported in the literature. Typical \( {\Delta G}_{sws}^{TOT} \) values reported for this surface material are − 29.3 (Van Oss 2005) and − 32 mJ m−2 (Simões et al. 2010). These contradictions may result from different nature, finishing, or cleaning treatment of the materials used (Simões et al. 2007).

The \( {\gamma}_s^{LW} \) parameter was statistically different (p < 0.05) between the surface materials, with the exception of PS-SS and Cu-PVC combinations. This parameter values ranged from 38.2 (for SS) to 40.2 mJ m−2 (for Cu). Comparing electron acceptor (\( {\gamma}_s^{+} \)) and donor (\( {\gamma}_s^{-} \)) parameters, it is possible to conclude that all surfaces are electron donors. Data obtained for \( {\gamma}_s^{+} \) and \( {\gamma}_s^{-} \) between the studied materials was statistically different (p < 0.05). The theory developed by Van Oss et al. (1987, 1988, 1989) predicts the non-existence of electron acceptor parameters, which corroborates with the \( {\gamma}_s^{+} \) values (close to zero). Looking at \( {\gamma}_s^{AB} \), the values determined ranged from 0.0 ± 0.0 to 13.8 ± 1.4 mJ m−2. In general, the \( {\gamma}_s^{AB} \) values were statistically different (p < 0.05) between the surface materials evaluated in this study, except for the PS-SS, PVC-SS, Cu-PMMA, and Cu-PS combinations.

Free energy of adhesion

In order to predict the ability of the selected microorganisms to adhere on the surface materials, the free energy of interaction, also known as free energy of adhesion, was calculated according to the thermodynamic approach represented in Eq. 1. As presented in Table 2, similar values of ∆Gadhesion were obtained for the selected microorganisms. Comparing the values obtained for each surface material, it is possible to observe that the total interfacial energy of the system was lower for Cu and G, with ∆Gadhesion < 0 mJ m−2. These results suggest that all studied microorganisms are expected to adhere on Cu and G surfaces. On the other hand, ∆Gadhesion was positive for SS, PMMA, PS, and PVC, in the decreasing order of free energy of adhesion. These results indicate that the interaction between microalgal/cyanobacterial cells and these surface materials is not thermodynamically favored and, consequently, it is not expected to occur. Comparing these surface materials, it is noticeable that adhesion may be thermodynamically less favorable on SS surfaces, due to the highest ∆Gadhesion values: values determined for each microorganism ranged between 19.7 ± 4.3 and 25.2 ± 4.3 mJ m−2.

Table 2 Free energy of adhesion (∆Gadhesion, mJ m−2) between the studied microorganisms and surface materials
Full size table

To better understand the relationship between the free energy of hydrophobic interaction and the free energy of adhesion, data from the \( {\Delta G}_{sws}^{TOT} \) for all surface materials were plotted against data from the ∆Gadhesion (for each microorganism). So, a good linear relationship was obtained, with coefficient of determination values (R2) of 0.994, 0.967, 0.985, and 0.982 for C. vulgaris, P. subcapitata, S. salina, and M. aeruginosa, respectively. This strong correlation is not surprising, as the thermodynamic approach is based on the surface physicochemical properties of the selected microorganisms and surface materials. Additionally, the positive relationship between free energy of hydrophobic interaction of the studied materials and free energy of adhesion shows that the larger the \( {\Delta G}_{sws}^{TOT} \) (more hydrophilic), the larger became the ∆Gadhesion.

Biofilm formation on the selected materials using different culture media

The density of attached cells varied with time, culture media, and surface among all microorganisms. Figure 1 presents the log CFUs per square centimeter determined on each of the surface materials using both culture media after 24, 72, and 168 h of incubation.

Fig. 1
figure 1

Number of adhered cells (log CFU cm−2) of C. vulgaris (a and b), S. salina (c and d), and M. aeruginosa (e and f) on the studied surface materials after 24, 72, and 168 h incubation in SE (a, c, and e) and OECD media (b, d, and f). The symbol * represents that no CFU was detected. Error bars correspond to the standard deviation of the mean determined for three independent experiments

Full size image

Comparing the biofilm formation ability for the different microorganisms, a significant difference (p < 0.05) was observed between the strains used in this study. C. vulgaris presented the highest ability to form biofilms, followed by M. aeruginosa, S. salina, and P. subcapitata (shown a lack of biofilm formation ability). In fact, P. subcapitata was unable to form biofilms on the selected materials; it was not possible to present the number of CFUs for this microalga. The inability of P. subcapitata for biofilm formation might be related to the high net zeta potential already reported for this microorganism (Gonçalves et al. 2015). According to Gonçalves et al. (2015), zeta potential determined for P. subcapitata was − 48.1 ± 0.9 mV, indicating that this microalga tends to be stable in the dispersed form.

For those microorganisms with biofilm formation ability, initial adhesion occurred after 24 h of incubation. Irving (2011), while studying the initial adhesion of C. vulgaris on various surfaces, obtained cell densities between 1.3 × 103 and 2.5 × 104 cells cm−2 after 7 days of growth on Petri dishes. Additionally, Sekar et al. (2004) investigated the attachment of C. vulgaris on pieces of several surface materials and found maximum attachment after 48 h of experiment. Additionally, Ozkan and Berberoglu (2013a), while measuring the strength of adhesion of C. vulgaris attached to different substrata, found maximum attachment after 10 h, which resulted in a density of 2.5 × 104 cells cm−2. Comparing the values obtained by the authors stated above with this study, it is possible to see that the present study reached higher biofilm densities for C. vulgaris, with maximum values of about 1.3 × 106 CFU cm−2. However, it is important to notice that in those studies they used flow chambers to induce biofilm formation, whereby the environmental and hydrodynamic conditions were different from those applied in this study. Comparing the different time points evaluated, it was possible to verify that, in general, the number of biofilm cells increased along time for all microorganisms and the highest biofilm density was reached after 7 days (168 h) of incubation. However, in a few cases, the number of CFUs decreased after 72 h and increased at 168 h of experiment, meaning that biofilm formation occurred within the first 24 h. The decreasing in the cell density after 72 h was not statistically significant (p < 0.05). These results may be related to the low EPS concentrations evidenced by SEM micrographs (Fig. 2c) and to the homogeneous distribution of microbial cells in the coupons (Fig. 2a), which results in a lower biofilm cohesion and helps to explain the low number of biofilm cells. Figure 2 shows representative SEM micrographs of SS coupons after 24 h of incubation with C. vulgaris. Such behaviour has already been reported in the literature (Menicucci Jr. 2010; Irving and Allen 2011; Kesaano and Sims 2014). This fact can also be related to nutrient depletion, which leads to cell migration from a surface with non-adsorbed nutrients to another location more favorable for their growth (Donlan 2002). In the study performed by Korber et al. (2003), the authors found that microorganisms can detach from a surface and remain planktonic or reattach after a few hours, depending on the microorganism. Furthermore, according to Horn et al. (2003), biofilm detachment can also be influenced by internal strength decrease, through hydrolysis of EPS. Although several studies on microbial attachment and detachment can be found in the literature, the mechanisms involved in detachment are not well established and further studies need to be done (Wilson et al. 2004). Other issues worth noting are (i) to know if the biofilm resists mechanical detachment, once initial cell adhesion occurs and the biofilm is established, and (ii) to compare the effect of physicochemical properties on both short-term and long-term experiments. Gross et al. (2016) found that both initial cell attachment and long-term attached growth are influenced by surface physicochemical properties of the microorganisms and the materials. Conversely, Irving (2011) found that although hydrophobicity could not be correlated with surface colonization, either in batch and continuous experiments, significant differences were observed in batch unlike continuous experiments, in which similar attachment patterns were found on both hydrophobic and hydrophilic surfaces. Furthermore, concerning biofilm growth, some authors stated that once initial colonization occurs on the surface material, the attachment of other cells can be much easier (Ozkan and Berberoglu 2013b; Katarzyna et al. 2015). The same researchers claimed that biofilm integrity is mainly governed by keeping the biofilm thickness at appropriate limits. Clearly, with increasing biofilm thickness, the bottom cell layers can become nutrient starved and light limited, leading to biofilm sloughing (Gross 2015). So, it is crucial to study biofilm growth kinetics over long-term experiments to find the most beneficial time for biofilm harvesting, in order to achieve higher biofilm density and thickness, while avoiding biofilm sloughing.

Fig. 2
figure 2

SEM micrographs of C. vulgaris attached on SS. a Magnification of 500×. b Magnification of × 2500. c Magnification of × 5000

Full size image

Regarding the surface materials, significant differences (p < 0.05) were observed in biofilm formation on the selected materials for all time points (for most cases). For the 24 h, the combinations PS-PVC, PS-Cu, SS-PVC, SS-Cu, PMMA-PVC, PMMA-Cu, G-Cu, and PVC-Cu presented statistically significant differences (p < 0.05). For the 72 h, significant differences (p < 0.05) were found between PS-Cu, SS-Cu, G-Cu, PVC-Cu, SS-PVC, SS-G, and SS-PMMA combinations. At 168 h, significant differences (p < 0.05) were observed between all the surface materials-Cu combinations.

In general, the degree of biofilm formation was found to follow the sequence SS > PS > G > PVC > PMMA> Cu (which corresponds to an average of adhered cells of 1.9 × 105 > 1.8 × 105 > 1.6 × 105 > 1.4 × 105 > 1.2 × 105 > 5.2 × 103 CFU cm−2). The degree of biofilm formation was based on which surface had the highest number of cells adhered (average of CFU cm−2) for the overall experiments, with both culture media and for all microorganisms. In the study performed by Sekar et al. (2004), the authors reported that surface colonization was higher on stainless steel, followed by titanium, PMMA, and glass with values of biofilm cells of about 8.0 × 103, 7.0 × 103, 6.8 × 103, and 3.8 × 103 cells cm−2, respectively.

Comparing the effect of different media compositions on biofilm formation, Fig. 1 shows that the number of CFUs determined when using the OECD test medium was higher than the one determined with SE. However, the differences were only statistically significant (p < 0.05) for 168 h. This increased biofilm formation ability under low nutrient concentrations explains the differences observed when using the selected culture media. Microalgae and cyanobacteria are mixotrophic, meaning that they can use both light and organic carbon as energetic source. Since the OECD test medium does not contain an organic carbon source, microalgae and cyanobacteria tend to be under stress conditions when grown on that medium. Previous studies (Fields et al. 2014; Kesaano et al. 2015) also proposed that environmental stress and nutrient deprivation enhances microalgal/cyanobacterial biofilm development, nutrient uptake, and lipid accumulation.

The influence of the microorganism, culture medium, surface material, and the combined effect of these variables on biofilm formation was assessed using a three-way-ANOVA (Table 3). Analysis of Table 3 shows that the combined effect of these factors presents a significant effect on biofilm formation (p < 0.05), especially for longer periods of incubation (72 and 168 h).

Table 3 Influence of the different variables evaluated in this study on cell adhesion obtained at different time periods
Full size table

Relationship between surface physicochemical properties and biofilm formation

Although the influence of surface physicochemical properties on bacterial adhesion has been extensively studied, only a few studies focused on a systematic analysis of microalgal/cyanobacterial biofilm formation on different materials (Sekar et al. 2004; Irving and Allen 2011; Ozkan and Berberoglu 2011; Sirmerova et al. 2013). Many researchers have reported that thermodynamic approaches alone cannot be used to surely predict bacterial attachment and biofilm formation (Morra and Cassinelli 1998; Li and Logan 2004; Chae et al. 2006; Simões et al. 2010). However, this fact is not well established for photosynthetic microorganisms and for the broad range of materials/surfaces used in this study. Comparison between the thermodynamic prediction of adhesion (Table 2) and the biofilm formation results present in Fig. 1 confirmed that microalgal and cyanobacterial biofilm formation cannot be predicted exclusively by thermodynamic approaches. For example, although ∆Gadhesion values determined for PMMA, PS, PVC, and SS assumed a positive value for both microalgae and cyanobacteria, proposing that the adhesion to these materials was thermodynamically unfavorable, the in vitro assays demonstrated the opposite.

On the other hand, data obtained from the analysis of physicochemical properties indicated that all microorganisms should have higher ability to adhere to Cu and G surfaces. In fact, G was one of the materials with more biofilm on its surface, as previously predicted by the thermodynamic approach. Nevertheless, Cu was the surface material presenting the lowest number of biofilm cells. Also, in most cases, there was no growth on Cu surfaces. Indeed, Cu has been reported as a toxic material to microalgae and cyanobacteria (Stauber and Florence 1987; Leale 1998; Lombardi et al. 2007; Jamers et al. 2013). Additionally, several authors (Van Leeuwen 1999; Costas and Lopez-Rodas 2006; Gregor et al. 2008; Hadjoudja et al. 2010) reported that M. aeruginosa is more sensitive to Cu than C. vulgaris, which corroborates data obtained; the adhesion experiments showed that there were no cyanobacterial cells adhered on Cu surfaces and, on the other hand, C. vulgaris was able to slightly attach to this material. The theoretical and in vitro adhesion of C. vulgaris to Cu, G, PMMA, PS, and SS has already been reported (Sekar et al. 2004; Irving and Allen 2011; Ozkan and Berberoglu 2013a, 2013b; Sirmerova et al. 2013). In these studies, the thermodynamic approach was not always in accordance with the adhesion assays, proving that more factors should be accounted in the prediction models.

Considering data obtained for surface physicochemical properties and comparing with those obtained for biofilm formation, a linear relationship was found. It is important to notice that for the linear association analysis, only data from 24-h biofilms using the OECD test medium were used, since the first day is critical for microalgal attachment and biofilm development (Di Pippo et al. 2009). The results obtained for Cu were not included, since, in general, there was no biofilm formation on this surface material. As it can be seen in Fig. 3, the correlation obtained was not good (R2 = 0.111). However, there is a negative relationship, since lower ∆Gadhesion values correspond, in general, to a higher number of biofilm cells. As it was described above, lower values of ∆Gadhesion are strongly associated with lower values of \( {\Delta G}_{sws}^{TOT} \) (hydrophobic surfaces). So, linking this correlation to the information presented in Fig. 3, it is possible to see that this relationship corroborates the hypothesis that cell adhesion tends to occur on hydrophobic surfaces rather than on hydrophilic ones, as proposed by several authors (Holland et al. 2004; Sekar et al. 2004; Stanley and Callow 2007; Li et al. 2010). In these studies, several microorganisms were used namely Phaeodactylum tricornutum, C. vulgaris, Nitzschia amphibia, Chroococcus minutus, Nitzschia closterium, Amphora coffeaeformis var. perpusilla, Craspedostauros australis, and Navicula perminuta. The discrepancies observed between the adhesion based on the thermodynamic theory and on biofilm formation in vitro assays can be explained by the non-consideration of environmental and microbiological parameters in the predictive approaches.

Fig. 3
figure 3

Linear regression between the number of colony forming units (CFU cm−2) and the free energy of adhesion (∆Gadhesion, mJ m−2)

Full size image

Despite the attempts of many researchers to describe microbial adhesion behaviour from individual physicochemical measurements of microorganisms and surfaces, the models proposed so far have significant limitations. Although the Van Oss (1995) approach is one of the most currently employed methods to predict microbial adhesion, it also has some limitations. Firstly, this model tends to ignore the hydrogen-bonding interaction of bacteria and surfaces with the liquid medium (Ista 2011). Secondly, it assumes smooth surfaces with a homogenous distribution of cells that are similar themselves (Ista 2011; Ista and López 2013). According to the DLVO theory, higher effective radius of particles leads to more repulsive electrostatic interactions. Accordingly, microalgae and cyanobacteria can reduce their effective radius of interaction with the surface material through some biological mechanisms, such as EPS production and the presence of extracellular motility appendages that can favor adhesion and biofilm formation (Doyle 2000; Sinde and Carballo 2000; Chaves 2004; Simões et al. 2007). In fact, Characklis (1990) reported that the transition from reversible to irreversible attachment is usually mediated by EPS, through the change from a weak interaction between microorganisms and surfaces to a permanent linkage. Furthermore, several studies suggested that the attachment of microalgae/cyanobacteria may be enhanced by the presence of bacteria on substrates by means of interspecific interactions (Bridier et al. 2014). Indeed, natural biofilms are mainly complex consortia of autotrophs, heterotrophs, and EPS (Hodoki 2005; Irving and Allen 2011; Shen et al. 2015). Besides, surface properties, such as roughness and texture, which have also been reported in the literature as key factors for microbial adhesion, are not included in the thermodynamic approach and could explain the differences observed between the prediction approach and the in vitro experiments (Sekar et al. 2004; Hodoki 2005; Shen et al. 2015). This lack of relationship between theoretical/thermodynamic and in vitro adhesion is well described for bacteria (Oliveira et al. 2001; Simões et al. 2007). However, there is still a lack of knowledge on the mechanisms involved in adhesion and biofilm formation of microalgae and cyanobacteria.

In conclusion, a comprehensive study has been performed to determine the surface physicochemical properties of selected microalgae, cyanobacteria, and surface materials and to understand the influence of these properties on biofilm formation. The influence of media composition was also assessed in this study. Free energy of hydrophobic interaction values showed that SS was hydrophilic, whereas Cu, G, PMMA, PS, and PVC were hydrophobic surfaces. Also, the selected microorganisms presented a hydrophilic character. Concerning the ability of the selected microorganisms to form biofilms on the referred materials, M. aeruginosa presented the highest biofilm formation ability, followed by S. salina and C. vulgaris, with maximum biofilm densities of 2.1 × 106, 1.3 × 106, and 2.9 × 105 CFU cm−2, respectively. P. subcapitata was the exception, as the results showed a lack of biofilm formation ability. Furthermore, the degree of biofilm formation was found to follow the order SS > PS > G > PVC > PMMA> Cu. Also, a significant difference (p < 0.05) between the number of biofilm cells when using the OECD test medium and SE was observed for 168 h, confirming that microbial adhesion may be enhanced by exposing microorganisms to nutrient stress conditions. Comparison between the thermodynamic theory and the experimental assays showed that adhesion may be underestimated when predicted exclusively by the thermodynamic approach and that its prediction depends on many other factors apart from surface physiochemical properties. Therefore, other factors should be taken into account, particularly the type and characteristics (presence/absence of extracellular appendages, production of EPS) of the microorganism, the bulk media composition, and the type of adhesion surface.