The purpose of this study was to investigate the feasibility of cultivating the biotechnologically important bacterium Streptomyces griseus in single-species and mixed-species biofilms using a tubular biofilm reactor (TBR). Streptomyces griseus biofilm development was found to be cyclical, starting with the initial adhesion and subsequent development of a visible biofilm after 24 h growth, followed by the complete detachment of the biofilm as a single mass, and ending with the re-colonisation of the tube. Fluorescence microscopy revealed that the filamentous structure of the biofilm was lost upon treatment with protease, but not DNase or metaperiodate, indicating that the extracellular polymeric substance is predominantly protein. When the biofilm was cultivated in conjunction with Bacillus amyloliquefaciens, no detachment was observed after 96 h, although once subjected to flow detachment. Electron microscopy confirmed the presence of both bacteria in the biofilm and revealed a network of fimbriae-like structures that were much less apparent in single-species biofilm and are likely to increase mechanical stability when developing in a TBR. This study presents the very first attempt in engineering S. griseus biofilms for continuous bioprocess applications.
Biofilms are microorganisms that grow attached to a surface and are characterised by the production of an extracellular matrix, increased tolerance to drugs and other xenobiotics, and improved intercellular communication. In clinical settings, biofilms are commonly found on medical implants and catheters; in industry, biofilms are problematic in pipes resulting in blockages and cause fouling and corrosion of crucial surfaces (Flemming, 2002; Hall-Stoodley et al., 2004). Biofilms play an important role in wastewater treatment, and their potential for degrading specific pollutants has been investigated (Misiak et al., 2011). In recent years, there has been a surge of interest in the application of single-species biofilms for biotechnology (Winn et al., 2012). Increased environmental robustness, self-immobilisation and their ease of integration into continuous flow bioreactors make biofilms ideal for production of fine chemicals at high productivities. For example, continuous production of (S)-styrene oxide from styrene was enabled by Pseudomonas sp. strain VLB120∆C cultivated in a tubular bioreactor, which had a volumetric productivity that exceeded that in a stirred tank reactor with the same strain and had a process duration of over 50 days (Gross et al., 2010). Biofilm catalysts employed in continuous processes also have the potential advantages of slower biomass turnover and consequently reduced biomass waste compared with equivalent batch systems.
Streptomyces spp. have a plethora of useful enzymes that can be applied to a range of potentially valuable biotransformations. One class of these is the cytochromes P450 that catalyses regio- and stereo-selective hydroxylation on unactivated carbon centres (Schulz et al., 2012). Streptomyces griseus is one of the best genetically characterised members of the genus and has many attributes that can be exploited for biotechnological applications. It is known to produce a number of bioactive molecules, including the antibiotic streptomycin, and analysis of the complete genome of S. griseus suggested that this strain has the capacity to produce over 34 secondary metabolites (Ohnishi et al., 2008). Furthermore, the potential of the intrinsic cytochrome P450 activity of S. griseus has been demonstrated to enable production of human-like drug metabolites of pharmaceuticals for subsequent toxicological evaluation (Alexandre et al., 2004; Bright et al., 2011). Despite the biotechnological importance of this bacterium, no studies have yet been conducted to investigate the potential of S. griseus biofilms in continuous flow biocatalysis. In this article, we describe the first investigations into the characteristics of S. griseus when cultured in a tubular biofilm reactor (TBR) and assess how co-cultivation with another bacterium might improve biofilm stability.
Materials and methods
Media and strains
Streptomyces strains were obtained from the German Collection of Microorganisms and Cell Cultures or were environmental isolates obtained from the UCD School of Biomolecular and Biomedical Science culture collection. Spore stocks were prepared from strains grown on solid ISP4 medium (BD Biosciences) and resuspended in 25% (v/v) glycerol and stored at −20 °C. Strains were grown at 28 °C in either soybean medium (5 g L−1 Bacto Soytone, 20 g L−1 glycerol, 5 g L−1 yeast extract, 5 g L−1 KH2PO4, pH 7.0), ISP2 (10 g L−1 malt extract, 4 g L−1 yeast extract, 4 g L−1 glucose, pH 7.2), tryptone soya broth (BD biosciences) or NMMP medium (van Keulen et al., 2003) supplemented with 20% (w/v) of either glucose or mannitol as carbon source. Planktonic cultures were grown in 50 mL screw capped vials fitted with a 2 cm spring and agitated at 180 r.p.m.
Bacillus amyloliquefaciens was obtained from the American Type Culture Collection (ATCC 23844) and grown in liquid LB broth (10 g L−1 tryptone, 10 g L−1 NaCl, 5 g L−1 yeast extract) or on solid LB and incubated at 28 °C. Planktonic cultures were agitated at 180 r.p.m.
Well plate assay
Streptomyces griseus ATCC 13273 was cultivated in polystyrene 12-well plates (Corning, Cellbind). Each well contained 2 mL of medium (ISP2, TSB or soybean) and was inoculated with 20 μL started culture. The well plates were incubated at 30 °C for up to 96 h, either statically or with gentle agitation (75 r.p.m.). Biofilm growth was assessed after removing the liquid, rinsing the wells with phosphate-buffered saline and staining with crystal violet; the absorbance was measured at 595 nm.
Tubular biofilm reactor
The TBR comprised of an inoculum reservoir, a medium reservoir fitted with a glass flow break, a peristaltic pump (Watson Marlow Sci Q400) connected to a 100-cm-long length of silicone tubing (8 mm OD, 5 mm ID; Fisher Scientific) and a spent medium reservoir (Fig. 1a). A planktonic starter culture was grown in soybean medium for 24 h and diluted 100-fold into the inoculum reservoir containing fresh medium. This inoculum was pumped through the reactor at 10 mL min−1 until the whole reactor was filled. The flow was switched off for 30 min to allow cells to adhere to the silicone tubing, and then, flow was initiated at 1 mL min−1 from the fresh medium reservoir. To test the effect of laminar flow on biofilm development, other reactors were inoculated either in the absence of laminar flow or the flow was initiated after 16 h following the formation of a thick biofilm. The TBR was operated for at least 26 h at 28 °C to observe the biofilm detachment cycle.
Miniature static TBR
To enable convenient screening of other Streptomyces spp. for biofilm production, a 20 cm length of sterile silicone tubing (8 mm OD, 5 mm ID; Fisher Scientific) was connected to a 6-mL luer lock syringe (Fisher Scientific). The syringe was used to draw appropriately inoculated soybean medium (prepared as for the TBR reactor) into the tubing, which was capped at the other end by an additional syringe. Adjustment of the two syringes was used to ensure no air bubbles remained inside the tubing. The tube was then incubated at 28 °C for 26–120 h (Fig. 1b).
Chemical degradation of the biofilm
Biofilms for chemical degradation experiments were grown in three-channel biofilm flow cells (BioCentrum-DTU, www.csm.bio.dtu.dk/Instrument%20Center/Resources/Biofilm%20Setup.aspx, see Fig. 1c). Flow cell channels were sealed by fixing a microscope coverslip (24 × 60 mm thickness range No. 1.5; Thermo Scientific) onto the flow cell with silicone rubber compound (RS components). The flow cell was connected to the flow system with silicone tubing (3 mm OD, 1 mm ID; Saint-Gobain Performance Plastics) and a peristaltic pump with a flow rate of 10 mL h−1 and sterilised by passing a 1% solution of Virkon (Antec International) through the channels for 15 min. The channels were then rinsed with PBS for an additional 15 min. To grow the biofilms, the channels were inoculated with an identical inoculum culture as prepared for the TBR reactor. The biofilms were allowed to develop under static conditions for 2–3 days before the growth medium was replaced with either H2O, DNase I (500 μg mL−1 in 5 mM MgCl2; Sigma), NaIO4 (40 mM in H2O; Sigma) or pronase (10 mg mL−1 in Tris-HCl, pH 7.5; Roche). Solutions of the chemicals were pumped through the channel for 10 min before being returned to static incubation at 30 °C.
Various Streptomyces cultures were prepared as described. Bacillus amyloliquefaciens was grown in a 50 mL screw cap vial in LB medium for 16 h at 28 °C. The starter cultures were diluted 100-fold into fresh soybean medium, and this co-culture was immediately used as the inoculum for either the continuous flow or miniature static TBR reactors. The biofilm was allowed to develop statically for 16–92 h.
For fluorescence microscopy, the biofilms in the flow cell were stained for 15 min with acridine orange (0.1%) and visualised with an Olympus BX51 microscope fitted with a BP460–490 excitation filter, a dichroic mirror (DM500) and a BA515–550 barrier filter.
Transmission electron microscopy
Samples for transmission electron microscopy (TEM) were first fixed in 2.5% glutaraldehyde in 0.1 M Sørensen's phosphate buffer [40.5 mL in 0.1 M Na2HPO4 (Riedel-deHaën AG, Seelze, Germany) and 9.5 mL in 0.1 M NaH2PO4 (Riedel-deHaën AG), pH 7.4] for a minimum of 2 h at room temperature and postfixed in 1% osmium tetroxide in Sørensen's phosphate buffer for 1 h at room temperature. Subsequently, the specimens were dehydrated in a graded ethanol series (30%, 50%, 70%, 90% and 100%). When dehydration was complete, samples were transferred from 100% ethanol to acetone, from acetone to a mixture of one part of acetone and one part of epoxy resin (24 g agar 100 resin), 9.5 g dodecenyl succinic anhydride (DDSA), 16.5 g methyl nadic anhydride (MNA) and 1 g 2,4,6- tri(dimethylaminoethyl)phenol (DMP-30) for 1 h. To complete the resin infiltration, the samples were placed in 100% resin at 37 °C for 2 h. Finally, samples were embedded in resin, placed at 60 °C for 24 h until polymerisation was complete.
For orientation purposes, sections from each sample were cut at 1 μm, stained with toluidine blue [consisted of toluidine blue (Agar Scientific, Essex, UK) and 1% sodium tetraborate powder, Na2B4O7.10H2O (May & Baker Ltd., Degenham, UK) in dH2O and examined by light microscopy (Leica DMLB, Leica Microsystems, Wetzlar, Germany)].
From these survey sections, areas of interest were identified and ultrathin (80 nm) sections were obtained using a Leica EM UC6 ultramicrotome (Leica Microsystems). These sections were collected on 200 mesh thin bar copper grids, stained with 2% uranyl acetate in H2O (20 min) and 3% lead citrate (5 min) and examined by TEM (Tecnai G2 20 TWIN; FEI Company, Hillsboro, OR) using an accelerating voltage of 80 kV or Tecnai G2 12 BioTWIN using an accelerating voltage of 120 kV).
Scanning electron microscopy
Samples for scanning electron microscopy (SEM) were processed the same way as the TEM samples until the last 100% ethanol step. From 100% ethanol, the samples were transferred to a mixture of 30% hexamethyldisilazane (HMDS) in acetone, from that mixture to a mixture of 60% HMDS in acetone, and then finally, the samples were placed in 100% HMDS. HMDS was allowed to evaporate at room temperature, and fully dried samples were then mounted onto metal stubs with double-sided carbon tape. Finally, a thin layer of gold was applied over the sample using an automated sputter coater (Agar Scientific). The samples were imaged using scanning electron microscope (Hitachi S-4300, Tokyo, Japan).
Results and discussion
Streptomyces griseus biofilm
Initial experiments to assess the biofilm-forming ability of S. griseus ATCC13273 were conducted using polystyrene well plates; however, no attachment was observed, and the bacterium grew at the air–liquid interface. A TBR was constructed (Fig. 1a) and inoculated with a 24-h-old planktonic culture of S. griseus ATCC 13273, grown in soybean medium, which was allowed a 30-minute attachment period before a flow of fresh medium was initiated at a rate of 1 mL min−1. Following incubation at 30 °C for 12 h, a thin layer of biomass was visible on the bottom of the tube. By 16 h, this initial layer had thickened to cover the lower 50% of the tubing (Fig. 2a). The nonmotile nature of Streptomyces spp. makes it more likely that a biofilm will develop on the base of the tube rather than over the entire surface as the bacteria are driven downwards by gravity.
The biofilm continued to grow and thicken for the next 10 h; however, 24–26 h following initial inoculation, the surface adhesion of the biofilm was lost. This was characterised by the entire biofilm becoming detached from the surface and curling into a cylindrical structure within the silicone tube (Fig. 2a). Under the laminar flow, this biofilm is pushed out of the reactor and can be collected in a long, continuous fragment (Fig. 2b). The detachment was found to be independent of the flow, as the same effect was observed if the biofilm was allowed to form entirely in the absence of flow or if the biofilm was allowed to develop for 16 h before flow was initiated. Increasing the flow rate to 2 mL min−1 also had no temporal effect on the biofilm detachment. The detachment was also observed in other complex media such as ISP2 or TSB and in the defined minimal NMMP medium with either glucose or mannitol as carbon source. Previous studies with Streptomyces coelicolor had highlighted that surface attachment is mediated by amyloid fimbriae that are activated in NMMP medium only when mannitol is used as a carbon source (de Jong et al., 2009).
Although all visible traces of the biofilm are removed at this point, a fresh tube-attached biofilm subsequently developed over the same time period as the initial biofilm, without further inoculation, and hence, some S. griseus cells remain attached to the tube. This fresh biofilm also became detached after 24 h of growth. This cycle of growth and detachment was seen to repeat at least three times inside the same TBR without any further inoculations. A detailed study of the surface of the biofilm by SEM confirmed that it consisted largely of interlocking hyphae forming a very thick mesh of cells (Fig. 2c), which accounts for the apparently strong cohesive properties.
Employing the TBR configuration shown in Fig. 1(b), other Streptomyces spp. were investigated for their biofilm-forming ability. All of the strains tested formed biofilm, and most showed signs of detachment within the timeframe of the experiment (Supporting Information, Fig. S1).
Biochemical analysis of the biofilm
To assess the chemical composition of the S. griseus ATCC 13273 biofilm and any extracellular matrix present, the biofilm was subjected to the standard triad of sodium metaperiodate, protease (pronase) and DNase treatment (Seidl et al., 2008) to determine whether carbohydrate, protein or extracellular DNA is present (Fig. 3). To enable convenient microscopic analysis of the degradation, biofilms were grown inside flow cells (Fig. 1c). Under these conditions, the biofilm grew much more slowly, possibly due to poor oxygenation within the channels; nevertheless, after 2 days, there was sufficient growth to allow treatment with the different reagents.
Treatment of the biofilm with DNase I for 48 h resulted in no change to the structure of the biofilm (Fig. 3b), suggesting that extracellular DNA is not a component of the extracellular polymeric substances (EPS). The addition of 40 mM sodium metaperiodate had mixed results; a number of filamentous structures remained visible even after 2 days of treatment, but evidence of biofilm breakdown could also be seen as truncated filaments and cell debris (Fig. 3c). The most marked change in biofilm structure occurred following pronase treatment (10 mg mL−1) as almost all the filamentous structure was lost and the biofilm was left as a collection of truncated filaments (Fig. 3d). Work by de Jong et al. (2009) on S. coelicolor inferred that adhesion of this bacterium was mediated by amyloid-like fimbriae that assemble along cellulose fibrils emerging from the surface of the cells. If S. griseus uses similar appendages, then proteases and the cellulose-degrading NaIO4 could be corroding these structures and therefore dispersing the biofilm cells. Proteases are also implicated in Streptomyces sporulation and may be breaking up the cells in addition to degrading putative fimbriae (Kim & Lee, 1995). The thick layers of filaments within the biofilm may act as a diffusion barrier to the NaIO4, which may explain the lack of complete dissociation with this reagent.
Streptomyces–Bacillus mixed culture biofilm
It has been previously shown with other bacteria that mixed culture biofilms have greater resistance to mechanical forces than single species (Simoes et al., 2009). Therefore, experiments were undertaken to improve the stability of the S. griseus biofilm by cultivating the bacterium in conjunction with another strain. Bacillus amyloliquefaciens was selected for the co-culture experiments as this organism is also known to produce a lactonase that might potentially disrupt sporulation of S. griseus by degrading the A-factor γ-butyrolactone signalling molecule (Yin et al., 2010). When inoculated with a mixed culture of S. griseus and B. amyloliquefaciens, it was observed that no detachment of the biofilm occurred inside a static TBR even after 96 h (Fig. S2), whereas S. griseus alone detached after 24 h when grown under the same conditions. A control experiment in which only B. amyloliquefaciens was cultured in the TBR resulted in a very weak biofilm coating the very bottom of the tube, and a similar co-inoculation of S. griseus with another bacterium (Escherichia coli) showed no similar effects on Streptomyces biofilm morphology or detachment. In the other Streptomyces strains that exhibited biofilm detachment, this was also delayed by co-culturing with B. amyloliquefaciens.
Scanning electron microscopy of the dual culture biofilm revealed that both bacteria are present and are closely associated with each other (Fig. 4a). TEM analysis of a cross-section of the single-species biofilm shows short hair-like projections on the surfaces of the cells that appear to make contact with adjacent cells (Fig. 4c). Other than these structures, extracellular matrix, which is characteristic of biofilms and can be observed in other bacterial biofilms with electron microscopy (Sriramulu et al., 2005; Kwiecinski et al., 2009; Tsoligkas et al., 2011), was not observed here. In fact, the filamentous structure of the biofilm produced by S. griseus more closely resembles that of fungal biofilms (Seidler et al., 2008; Amadio et al., 2013). In S. coelicolor, fimbriae composed of bundled amyloid fibrils of chaplin protein are responsible for attachment of the bacterium to surfaces. Similar fimbriae might be involved in the biofilm of S. griseus, which would be consistent with the dispersal of biofilm in the presence of pronase as observed in the flow cell experiments. Furthermore, TEM images of cross-sections of the dual species biofilm revealed a more extensive network of the putative fimbriae than was observed in the single-species S. griseus biofilm (Fig. 4b), which would account for the improved stability. This more robust S. griseus/B. amyloliquefaciens biofilm still detached when a flow of fresh medium was initiated (1 mL min−1), suggesting that the adhesive property of the biofilm is still not strong enough to withstand moderate shear forces.
Although a small number of studies on streptomycete biofilms has been conducted (Kim & Kim, 2004; Morales et al., 2007; de Jong et al., 2009), none has been studied in a continuous flow biofilm reactor. In this article, we have described for the first time the characteristics of the biotechnologically important bacterium S. griseus when cultivated as a biofilm in a tubular reactor. The biofilm that forms is relatively unstable and detaches after 24 h; however, without further inoculation, fresh biofilm regrows on the tube. The cohesive strength of the biofilm is evident as the biofilm detaches as a single piece of biomass, and this is accounted for by the dense network of filaments observed by SEM. Although no extracellular matrix was apparent in the electron micrographs, pronase, and to a lesser degree metaperiodate, caused the biofilm to disintegrate, suggesting that these polymers form the EPS. Co-cultivation of S. griseus with B. amyloliquefaciens resulted in a more stable biofilm that could be at least partially explained by the observation by TEM of a much more extensive network of putative fimbriae compared with single-species biofilm. It is likely that these fimbriae provide improved mechanical strength to the biofilm.
Streptomyces griseus produces important secondary metabolites and catalyses biotechnologically relevant reactions, and these attributes might be further exploited by employing biofilms. The findings reported here suggest that in a submerged environment S. griseus is capable of forming thick surface-attached biofilms. However, spontaneous surface detachment occurs, and controlling this detachment is the key to forming a stable biofilm, thus enabling continuous flow applications.
This work was supported by Science Foundation Ireland under Grant 11/TIDA/B2007. The authors thank Dimitri Scholz and Tiina O'Neil, UCD Conway Institute, for assistance with the electron microscopy. The authors confirm no financial interest or benefit arising from the research.