Introduction

The ability to form biofilms is considered the most prevalent means of microorganisms to persist in nature, enabling microbes to withstand a broad variety of environmental fluctuations such as temperature and pH changes, nutrient availability and the presence of toxins (Costerton et al., 1995; Lopez et al., 2010). Biofilms have been extensively studied in members of the bacteria as their presence can promote several persistent and chronic infections (Costerton et al., 1999). In contrast, it is rather recent that environmental biofilms built up of archaeal and bacterial species are being examined, mainly due to their relevance in biogeochemical cycling of essential elements (Shock et al., 2005; Justice et al., 2012).

Archaea form biofilms within many microbial ecosystems such as acid mine drainage sites, seafloor sediments or acidic hot springs mats (Baker and Banfield, 2003; Orcutt et al., 2011); Kozubal et al., 2012). Initial description of the archaeal biofilms were reported in the euryarchaeota Archaeoglobus fulgidus (Lapaglia and Hartzell, 1997) and in the bi-species biofilm of Pyrococcus furiosus and Methanopyrus kandlerii (Schopf et al., 2008). Ferroplasma acidarmanus displayed a multilayered biofilm and proteomic studies revealed upregulation of proteins involved in the adaptation to anoxia indicating existence of anaerobic zones in the multilayered biofilms (Baker-Austin et al., 2010). Morphologies of different haloarchaeal biofilms ranged from carpet-like to multi-layered biofilms containing micro- and macro-colonies as well as biofilms characterized by large aggregates of cells able to attach to abiotic surfaces (Fröls et al., 2012).

Three crenarchaea S. acidocaldarius, S. solfataricus and S. tokodaii displayed very diverse biofilm morphologies: either simple carpet-like structures in S. solfataricus or highly dense tower-like structures in S. acidocaldarius communities (Koerdt et al., 2010). Proteomic studies conducted on biofilms grown as static biofilms of S. acidocaldarius, S. solfataricus and S. tokodaii showed that only seven changes were shared across the three strains (Koerdt et al., 2011). One of the most striking common response genes included the putative Lrs14-like transcriptional regulators indicating their possible role as regulatory factors during biofilm development of Sulfolobus spp.

Although an Lrs14 protein-encoding gene was previously identified and isolated from S. solfataricus (Napoli et al., 1999), its physiological target genes are unknown. S. solfataricus Lrs14 was found to be negatively auto-regulated and accumulated in late growth phases (Napoli et al., 1999). In contrast to other known metal-dependent regulators inhibition occurred in a ligand-independent manner (Bell and Jackson 2001).

In this study, we investigated the effect of the deletion of six Lrs14 transcriptional regulators during biofilm development of S. acidocaldarius. Mutational analysis combined with both phenotypic and in vitro characterization of the six homologous Lrs14 proteins encoded by S. acidocaldarius revealed that three of them significantly influenced either biofilm formation or cell motility. For one homolog, saci0446, it was shown that its absence strongly impaired cell motility and promoted extracellular polymeric substance (EPS) overproduction, thus leading to an enhanced biofilm formation. This report provides for the first time insights into transcriptional regulation of archaeal biofilm development.

Materials and methods

S. acidocaldarius strains and growth conditions

S. acidocaldarius MW001 (Wagner et al., 2012) and all in-frame marker-less deletion mutants were aerobically grown at 76 °C in Brock media (Brock et al., 1972), pH 3 and supplemented with 0.1% (w/v) N-Z-amine and 10 mg ml−1 uracil. Uracil was not added to the media for cultivation of S. acidocaldarius MW001pyrEF+ and pyrEF disruption mutants. For protein overproduction experiments in S. acidocaldarius, 0.4% (w/v) maltose was added to the media to induce expression. Growth progression was monitored by the measurement of the optical density at 600 nm (OD600). All S. acidocaldarius deletion strains are described in Table 1.

Table 1 Strains and plasmids
Full size table

Methods for protein alignments, the construction of plasmids for deletion mutants and the genetic manipulation of S. acidocaldarius are described in the Supplementary Material. Oligonucleotides employed for these procedures are listed in Supplementary Table 1.

Microtitre plate assays

A microtitre plate assay using polystyrol 96-well tissue culture plates (flat bottom cell+, Sarstedt, Nuembrecht, Germany) adapted to high temperature as developed by Koerdt et al., 2010 was performed. After 2 days incubation, microtitre plates were cooled down to room temperature and the efficiency of biofilm formation was calculated by the correlation of the measured crystal violet absorbance of attached cells (OD570) and growth of planktonic cells (OD600). At least eight plates were used for both deletion and reference strains. The results were represented as percentage of biofilm formation of the S. acidocaldarius Lrs14 deletion mutant strains relative to either S. acidocaldarius MW001 or MW001pyrEF+ reference strains.

Biofilm culturing and confocal laser scanning microscopy (CLSM) analysis

Static biofilm cultures of S. acidocaldarius strains were grown in small Petri dishes (μ-dishes, 35 mm, Ibidi, Martinsried, Germany) in Brock media supplemented with 0.1% (w/v) N-Z amine and 10 mg ml−1 of uracil when necessary. For overexpressing strains 0.2% (w/v) maltose was added. All strains were inoculated at 0.01 OD600 and two biological replicates of each strain were grown for 3 days at 75 °C. The medium was carefully exchanged every 24 h to ensure aerobic growth conditions and nutrient replenishment. Petri dishes were put in a specially designed metal box (25 cm L × 20 cm W × 20 cm D) filled with ∼500 ml of water in the bottom to minimize evaporation of the media, as described by Koerdt et al. (2010). Biofilms were imaged as described in Koerdt et al. (2012).

To evaluate cell surface coverage of the biofilms, pictures of the bottom layer were taken using a differential interference contrast objective. Twelve images at different microscopy fields were recorded. By using Adobe Photoshop CS2 software differential interference contrast pictures were converted into black/white in order to calculate number of pixels/area, thus representing the percentage surface coverage. Cell surface coverage determinations were performed in three biological replicates.

Swimming motility on semi-solid gelrite plates

Swimming motility on plates was analyzed on semi-solid gelrite plates consisting of 0.15% gelrite supplemented with 0.001% (w/v) NZ-amine and 0.4% (w/v) maltose when necessary. Cells grown in standard Brock medium were harvested during exponential growth and used to inoculate plates at a cell density of 107 cells per ml. Plates were incubated for 7 days in a humid chamber at 75 °C. Swimming behavior of the different S. acidocaldarius strains was analyzed by measuring the swimming radius.

RNA isolation and quantitative reverse transcription-PCR (qRT-PCR)

Total RNA samples were isolated from 10 ml of exponentially growing shaking culture (OD600=0.2), 10 ml of shaking culture at stationary phase (OD600=0.4) and 40 ml of 3 days mature biofilm culture. To preserve RNA integrity, biofilm-containing Petri dishes were cooled down on ice before isolation and shaking cultures were immediately harvested by centrifugation at 4 °C. TRIzol reagent (Invitrogen, Carlsbad, CA, USA) was used for total RNA isolation following manufacturer’s instructions. The preparation of the complementary DNA and the qRT-PCR were performed as described in Lassak et al., 2012. Cq values of each transcript of interest were standardized to the Cq value of the housekeeping gene saci0574 (secY; Van Der Sluis et al., 2006). Quantitative PCR reactions with DNA-free RNA as template were performed as control. Primers used for quantitative PCR reactions are listed in Supplementary Table 1. At least three biological replicates of each assessed condition and two technical replicates per quantitative PCR reaction were performed.

EPS isolation and quantization

For EPS extraction, S. acidicaldarius strains MW001, MW251 and MW264 were grown as static biofilm cultures in 70 ml Brock media using 150 mm diameter polystyrene Petri dishes (Sarstedt). Brock medium was supplemented with 0.4% maltose when necessary to induce protein expression. After 3 days of incubation at 76 °C, biofilms of each strain were scraped off from four Petri dishes and washed three times with 10 ml phosphate buffer (6 mM, pH 7). EPS were then isolated using the cation exchange resign Dowex (Sigma-Aldrich, Munich, Germany) as described previously for the isolation of EPS from S. solfataricus biofilms (Koerdt et al., 2012). Carbohydrates ⩾3.5 kDa were considered as high molecular weight.

Expression, purification and activity tests with saci0446 are described in the Supplementary Material. Oligonucleotides employed for the construction of expression plasmids are listed in Supplementary Table 1.

Results

Lrs14-like proteins are an archaea-specific class of transcriptional regulators

BLAST searches using Saci1223, a putative Lrs14-like protein from S. acidocaldarius, as query revealed that hits showing ⩾80% of query sequence coverage only matched to archaeal amino-acid sequences, corresponding to 58 crenarchaeal sequences (expected value ⩽3e-04) and 19 euryarchaeal homologous sequences (expected value ⩽8e-04). Only one bacterial sequence was retrieved when using Saci1223, which was a putative transcriptional regulator (YP_004180336.1; expected value of 7e-04), encoded by the moderate thermophilic bacterium Isosphaera pallida, which shared 30% of identity with Saci1223. In addition, the BLASTP analysis determined that S. acidocaldarius genome encodes for six homologous Lrs14-like proteins: Saci1223, Saci0102, Saci0133, Saci0446, Saci1219 and Saci1242. These homologous proteins share identities from 34% to 39%.

Lrs14 proteins were formerly described as a bacterium-type transcriptional regulator related to the Lrp/AsnC (leucine-responsive regulatory protein) family of transcriptional regulators (Napoli et al., 1999). Sequence alignment analysis together with secondary structure predictions revealed that the Lrs14 proteins lacked the C-terminal RAM domain (a βαββαβ-fold motif), which is the distinctive ligand-binding domain of Lrp/AsnC-like proteins (Peeters and Charlier, 2010; Supplementary Figure 1A). Moreover, a neighbor joining distance analysis of archaeal Lrs14 and Lrp/AsnC regulators demonstrated that each subset of transcriptional regulators clustered in two well-defined clades, further supporting the distinctiveness between Lrs14-like and Lrp/AsnC as two divergent classes of regulators (Supplementary Figure 1B). Taking our analysis into account, we propose to categorize Lrs14-like proteins as a distinct type of archaea-specific class of transcriptional regulators.

Biofilm-associated transcriptional profile of S. acidocaldarius lrs14 genes

Previously, two of the six lrs14 genes (saci1223 and saci1242) present in the S. acidocaldarius genome were found to be upregulated in biofilm-associated cell populations when compared with their planktonic counterparts (Koerdt et al., 2011). To obtain the expression profiles of all six S. acidocaldarius lrs14 genes, RNA was isolated from cells grown as 3-day-old biofilms or as planktonic shaking culture either in the exponential and late stationary phase. The expression of each lrs14 gene was then determined as relative to the transcript levels found at the exponential growth phase. As shown in Figure 1, transcript levels of all six Lrs14-encoding open reading frames were increased both in the stationary growth phase and during biofilm growth. The expression of all six lrs14 genes was enhanced at least twofold in the stationary phase (Figure 1). The S. solfataricus Lrs14 (SSO1101) also showed higher transcript levels at late growth stages (Napoli et al., 1999). The most noteworthy changes were observed for saci0446, saci1223 and saci1242, which were highly induced in S. acidocaldarius biofilms (9, 7 and 32-fold changes, respectively). The same gene expression profile was observed when comparing the biofilm-associated cell population versus its planktonic cell population counterpart (Supplementary Figure 2). These gene expression patterns implied a potential role of Lrs14 proteins Saci0446, Saci1223 and Saci1242 during biofilm development.

Figure 1
figure 1

Expression profile of S. acidocaldarius lrs14 genes during both biofilm and planktonic growth. Total RNA isolated from S. acidocaldarius MW001 and MW001pyrEF+ (reference strains) grown either as biofilms or as planktonic cultures were used for complementary DNA (cDNA) synthesis. qRT-PCR analysis was performed using specific primers for each Lrs14-encoding ORF (shown underneath the plot). Relative transcript expression levels of each gene were normalized to the internal control gene secY. The values reflect the fold change in gene expression compared with cDNA prepared from exponential grown planktonic references strains cells, which is designated as baseline. The means and standard deviations of three biological replicates are shown.

Full size image

Biofilm formation of S. acidocaldarius lrs14 single deletion mutants

To understand the in vivo function of the Lrs14 proteins, single deletion mutants of all six lrs14 putative genes were constructed in S. acidocaldarius. Using the marker-less mutant method in the uracil auxotrophic S. acidocaldarius mutant MW001 as reference strain (Wagner et al., 2012), in-frame deletion mutants were obtained for saci0102, saci0446 and saci1242. As this strategy was not successful for the construction of saci0133, saci1223 and saci1219, these ORFs were deleted by a single homologous recombination step disrupting each gene via the insertion of the pyrEF selection cassette. The identity of each mutant strain was confirmed by PCR amplification of the appropriate genomic region and the subsequent sequencing (Supplementary Figure 3).

Analysis of the growth curves in shaking cultures revealed no obvious difference in the growth kinetic of the deletion mutants compared with the respective reference strains, MW001 (for in-frame deletion mutants) or MW001+pyrEF (for pyrEF disruption mutants; Supplementary Figure 4). Only slightly lower cell densities at the stationary phase were observed for Δsaci0446, Δsaci1242, Δsaci0133 and Δsaci1219 deletion strains (Supplementary Figure 4). Interestingly, the diminished biomass shown by the Δsaci0133 and Δsaci1219 deletion strains at late growth stages correlated with the higher transcript levels found in the stationary growth (Figure 1). In addition, no morphological defects could be observed when Lrs14 deletion mutant strains cells were subjected to optical microscopic analysis (data not shown).

The ability of Lrs14 deletion strains to form static biofilms was assessed by means of a microtiter plate assay adapted to high temperatures (Koerdt et al., 2010). After 3 days of static biofilm formation, three out the six deletion mutants showed significant alterations (Figure 2). Although Δsaci1223 showed a 60% decrease in biofilm formation, in the Δsaci0446 mutant 40% more biofilm was formed compared with the reference strain. In addition, biofilm formation by the Δsaci1242 mutant was reduced by 20% when compared with the reference strain (Figure 2). Mutant strains Δsaci0102, Δsaci0133 and Δsaci1219 revealed no significant differences in biofilm formation (Figure 2). Taken together, these results strongly suggested that Lrs14 transcriptional regulators Saci1223, Saci0446 and Saci1242 have a role during development of S. acidocaldarius biofilm communities.

Figure 2
figure 2

Biofilm formation of the S. acidocaldarius lrs14 deletion mutants by microtitre plate assays. Biofilm formation of each strain was calculated by the correlation of the measured crystal violet absorbance of attached cells (OD570) and growth of planktonic cells (OD600) to emphasize the amount of cells in a sessile lifestyle. The graph shows biofilm formation as relative to the wild-type strain MW001, which represented 100%. Each point and standard deviation is the mean of at least eight plates per condition. *Significant P⩽0.05, **highly significant P⩽0.01.

Full size image

Comparative analysis of biofilm architectures formed by S. acidocaldarius Lrs14 deletion mutants

In order to examine the morphologies of the biofilms formed by the Lrs14 deletion mutants, the reference strains (MW001 and MW001pyrEF+) and the Lrs14 deletion mutants were grown as static biofilms for 3 days. Although 4′-6-diamidino-2-phenylindole was used for visualization of cells within the biofilms, the presence of extracellular polysaccharide residues was detected using fluorescently labeled lectins that specifically bound to mannose/glucose (ConA) and galactosyl sugar residues (IB4). In Figure 3a, 4′-6-diamidino-2-phenylindole signal images (left column) or the overlay images of the three fluorescent signals are depicted (right column). Moreover, the surface coverage of each biofilm formed by the investigated strains was determined (Figure 3b).

Figure 3
figure 3

CLSM analysis of biofilm formed by the S. acidocaldarius lrs14 deletion mutants. (a) Three-day-old biofilms were subjeted to CLSM. The blue channel is the 4′-6-diamidino-2-phenylindole (DAPI) staining. The green channel represents the fluorescently labeled lectin ConA that binds to glucose and mannose residues. The lectin IB4 able to bind to a-galactosyl residues is shown in yellow. Overlay images of all three channels are shown. (b) Differential interference contrast (DIC) pictures (left panel) were taken from the bottom layer of biofilms and converted into black/white (B/W; right panel) to calculate the surface coverage. Numbers represent the percentage of surface coverage for each mutant strain.

Full size image

As described by Henche et al. (2011), after 3 days of growth reference strain MW001 showed a confluent dense biofilm, displaying an EPS pattern in which the ConA (mannose/glucose) signal was dominant (Figure 3). Biofilm communities formed by MW001pyrEF+ showed no differences in comparison with MW001 (Supplementary Figure 5). Deletion mutants Δsaci0102 and Δsaci0133 showed biofilm architectures resembling the reference strain biofilm phenotype. Only slight differences were distinguishable as moderate lower cell density or less EPS production, leading to a decrease in biofilm heights when compared with the wild type (Figure 3). Δsaci1219 displayed a rather uneven biofilm phenotype in comparison with the reference strain (Figure 3). As expected from the microtitre plate assays, strains Δsaci1223, Δsaci0446 and Δsaci1242 showed pronounced morphological differences when forming biofilms. A poorly cell-colonized surface was observed for the biofilms of Δsaci1223 and Δsaci1242 deletion strains as observed by 4′-6-diamidino-2-phenylindole signal and the percentage of surface coverage of each strain (Figure 3). In addition, although Δsaci1223 produced only small amounts of EPS, Δsaci1242 displayed a particular cloud-like EPS pattern unevenly distributed on top of the biofilm (Figure 3). In contrast, the deletion of saci0446 resulted in a densely packed biofilm structure. Although Δsaci0446 biofilm phenotype resembled the one displayed by the reference strain in terms of cell density, biofilm height and cell surface coverage, EPS production was notably increased by this mutant (see below), as clouds of mannose/glucose-rich EPS regularly distributed on top of the biofilm was visualized (Figure 3a).

The deletion of saci1223, saci1242 and saci0446 had a clear impact on formation and structures of S. acidocaldarius biofilm communities. In the deletion strains Δsaci1223, Δsaci1242 biofilm formation was impaired whereas the Δsaci0446 deletion strain exhibited a highly increased production of EPS leading to an enhanced biofilm formation.

Lrs14 protein Saci0446 controls cell motility of S. acidocaldarius

S. acidocaldarius exhibits three distinct type IV pili-like structures on its surface: (i) the archaellum, (ii) the ultraviolet-induced pili and (iii) the adhesive pili aap. All three cell surface appendages played a role in surfaces colonization and the interplay of the three structures is important for the MW001 biofilm phenotype (Henche et al., 2011). Therefore, it was important to determine whether the deletion of saci1223, saci1242 or saci0446 led to an altered synthesis of any of these cell surface appendages and therefore to the observed Lrs14 deletion mutant biofilm phenotypes.

Using qRT-PCR, the expression levels of gene-encoding components, which are essential for the assembly of each cell surface appendage (Henche et al., 2011), were determined in 3-day-old biofilm communities of Δsaci1223, Δsaci1242 and Δsaci0446, respectively. We determined that none of the tested cell surface appendage genes were differentially expressed in the Δsaci1242 deletion strain as their transcript levels were found to not be significantly altered (considering a threshold of ⩾±2-fold changes; Figure 4a). A similar expression pattern was observed for the Δsaci1223 mutant, where only the expression of upsA and upsE (ultraviolet-induced structural component-encoding genes) were determined as downregulated (2.8- and 2.2-fold changes, respectively; Figure 4a).

Figure 4
figure 4

Effect of S. acidocaldarius lrs14 deletion mutants on cell motility. (a) qRT-PCR experiments to determine gene expression of components of the archaellum (flaB and flaX), aap pili (aapA and aapF) and ultraviolet-induced pili (upsA and upsE) in the deletion strains Δsaci0446 (black bars), Δsaci1242 (gray bars) and Δsaci1223 (white bars) during biofilm growth. Relative transcript expression levels of each target gene were normalized to the internal control gene secY. The values reflect the fold change in expression compared with the reference strain MW001, which is designated as baseline. The means and s.d. of three biological replicates are shown. (b, c) FlaB expression levels were detected in Lrs14 deletion mutants by immunoblotting with specific antibodies before (−) and after (+) induction via tryptone starvation. (d) Motility assay of Lrs14 deletion mutants in comparrison to the reference strain MW001 and MW001pyrEF+. The non-motile strain ΔfalJ was included as a negative control. Trans-complemented strains of Δsaci0446 are shown.

Full size image

In contrast, transcript levels of flaB and flaX (essential components of the archaellum) were downregulated in Δsaci0446 biofilms (3.4- and 2.0-fold changes, respectively; Figure 4a), whereas transcript levels of aapA (3.4-fold changes), one of the two pilins forming the adhesive pilus, were increased (Figure 4a). Interestingly, this transcription profile resembled the one described for the S. acidocaldarius ΔflaJ strain (a non-archaellated strain) as it showed an increase of aapA and aapB transcript levels when grown as biofilms (Henche et al., 2011). The overexpression of aapA in the Δsaci0446 deletion strain might contribute its ability to form a more dense and stable biofilm because of the role of this adhesive appendage in both surface attachment and the establishment of cell-to-cell connections (Henche et al., 2011).

To test the presumably regulatory role of Saci0446 in archaellum expression, protein levels of FlaB, the structural protein of the archaellum, were determined in Δsaci0446, Δsaci1223 and Δsaci1242 in tryptone-starved cells, as it has been shown previously that upon starvation biosynthesis of all archaellum subunits is initiated (Lassak et al., 2012). Expression levels of FlaB were the same in Δsaci1223 and Δsaci1242 deletion strains compared with the reference strain (Figure 4b), whereas the accumulation of FlaB was hardly visible in the Δsaci0446 deletion strain (Figure 4b). Moreover, FlaB wild-type proteins levels could be restored in trans-complemented Δsaci0446 deletion strains by using a plasmid harboring the saci0446 coding sequence including either its own promoter (Δsaci0446+p_saci0446) or a maltose inducible promoter (Δsaci0446+pmal_saci0446; Figure 4c).

To confirm that the deletion mutant Δsaci0446 lacks archaella, its ability to swim on semi-solid gelrite plates containing reduced amounts of tryptone (0.005%) was tested. Motility of the Δsaci0446 deletion strain was decreased to levels, which were comparable with the non-motile S. acidocaldarius mutant strain ΔflaJ (Lassak et al., 2012; Figure 4d). Motility of Δsaci0446 cells could be trans-complemented with the gene under control of its own or the malE promoter (Figure 4d). In addition, the Δsaci1242 deletion strain showed a swimming radius comparable to the reference MW001, whereas Δsaci1223 motility was slightly impaired (Figure 4c). In conclusion, Saci0446 is involved in regulating motility of S. acidocaldarius, most likely by controlling expression of archaella components.

EPS characterization of biofilms formed by Δsaci0446

As already described, Δsaci0446 deletion strain biofilms exhibited a prominent EPS production pattern (Figure 3), implying a regulatory role of Saci0446 in the control of EPS biosynthesis. To test this assumption, trans-complemented Δsaci0446 deletion strains were subjected to CLSM analysis. The presence of saci0446 under the control of its promoter (Δsaci0446+p_saci0446) indeed reduced the production of EPS (Figure 5a), and complementation with saci0446 under overexpression conditions led to even lower EPS levels than apparent in the reference stain biofilms (Figure 5a). In addition, biofilms formed by the Δsaci0446+pmal_saci0446 strain displayed higher cell densities when compared with the reference strain (Figure 5a).

Figure 5
figure 5

EPS analysis of the S. acidocaldarius deletion strain saci0446. (a) Three-day-old biofilms of trans-complemented Δsaci0446+p_saci0446 and Δsaci0446+pmal_saci0446 strains were analysed by CLSM and compared with the reference strain MW001 and Δsaci0446 strains. Biofilm cells were stained using 4′-6-diamidino-2-phenylindole (blue), ConA (green) and IB4 (yellow). The overlay images of all three channels are shown. Scale bar=20 μm. (b) EPS were isolated from 3-day-old biofilms of MW001, Δsaci0446, Δsaci0446+pmal_saci0446 and MW001 suplemented with 0.4% (w/v) of maltose. Carbohydrate concentrations were determined from both, non-dialyzed EPS (total EPS, gray bars) and dialyzed EPS extracts (3.5 kDa; high-molecular-weight EPS, black bars). (c) Protein concentrations of the same non-dialyzed and dialyzed EPS.The means and s.d. of three biological replicates are shown for b and c.

Full size image

To determine differences in composition and quantity of EPS components during biofilm formation, MW001, Δsaci0446 and Δsaci0446-pmal_saci0446 strains were grown as static biofilms and EPS were isolated using a cation exchange resin (Dowex). Subsequently, the carbohydrate and protein contents of the non-dialyzed cell-free EPS fractions (that is, total amount of extracellular carbohydrates and proteins including low-molecular-weight substances) and the dialyzed cell-free EPS fractions were determined (Figures 5b and c). The EPS of the reference strain MW001 contained a total carbohydrate concentration of 7.4 fg per cell. Most of the measured carbohydrates in MW001 were of low molecular weight (⩽3.5 kDa) and were thus removed by dialysis resulting in an EPS carbohydrate concentration of 1.8 fg per cell (Figure 5b). The mutant strain Δsaci0446 displayed a fivefold higher carbohydrate production per cell (38.4 fg per cell), compared with the MW001 strain (Figure 5b). The carbohydrate levels of the EPS could be restored to wild-type levels in the trans-complemented Δsaci0446+pmal_saci0446 recombinant strain (Figure 5b). To exclude an effect on EPS production by the presence of maltose that was needed for the expression of saci0446, the carbohydrate and protein content of the EPS content was also determined in MW001 cells grown in the presence of maltose. These controls did not show any changes when compared with MW001 grown in the absence of maltose (Figures 5b and c). In addition, protein quantification revealed a similar trend as observed for the carbohydrate content of the EPS in all tested strains (Figure 5c). The highest secretion of proteins was found in the Δsaci0446 deletion strain, showing 12.9 fg per cell, with 76.8% being proteins of high molecular weight (Figure 5c). This analysis indicated that the extracellular biofilm matrix of the Δsaci0446 deletion strain was highly enriched in both polysaccharides and proteins.

To date, no genes involved in EPS biosynthesis in archaea are known. However, the S. acidocaldarius genome contains a gene locus (saci1904–1927) encoding 11 glycosyltransferases and other genes coding for enzymes involved in the modification and polymerization of sugars. Within this gene cluster, we focused on putative gene products, which share similarities with enzymes involved in bacterial exopolysaccharide biosynthesis and secretion (Figure 6a). Therefore, the differential expression profiles of those genes in Δsaci0446 biofilm-associated cells in comparison with the reference strain MW001 were determined. Although transcript levels of saci1908 were increased in the Δsaci0446 deletion strain, transcription of saci1909 was downregulated (Figure 6b). saci1908 encodes a putative membrane protein exhibiting 15 transmembrane segments and saci1909 encodes a putative glycosyltransferase sharing 28.8% sequence identity with a dolichyl-phosphate mannose synthase. The functions of these two gene products are still uncharacterized in S. acidocaldarius.

Figure 6
figure 6

Analisys of S. acidocaldarius putative genes involved in EPS production. (a) EPS-related gene cluster of S. acidocaldarius. White arrows represent ORF with homology to glycosyl transferases encoding genes, whereas gray arrows correspond to ORFs which gene products share homology with proteins involved in bacterial EPS biosynthetic pathways. Best blast hits are indicated underneath. (b) Diferential gene expression of EPS-related genes in the deletion strains Δsaci0446 (black bars) during biofilm growth. The values reflect the fold change in expression compared with the reference strain MW001, which is designated as baseline. Relative transcript expression levels of each target gene were normalized to the internal control gene secY. (c) CLSM analysis of 3-day-old biofilm cultures of the double and single deletion mutants: Δsaci0446-saci1908, Δsaci1908, Δsaci0446-saci1909 and Δsaci1909. Biofilm cells were stained using 4′-6-diamidino-2-phenylindole (blue), ConA (green) and IB4 (yellow). The overlay images of all three channels are shown. Scale bar=20 μm.

Full size image

In order to shed light into the potential role of saci1908 in S. acidocaldarius exopolysaccharide biosynthesis, double mutants using Δsaci0446 as a background strain were generated for both saci1908 (Δsaci0446-saci1908) and saci1909 (Δsaci0446-saci1909). The distinctive Δsaci0446 biofilm phenotype was reverted in Δsaci0446-saci1908 and resembled reference strain MW001 biofilms (Figure 6c). In contrast to this, the deletion of saci1909 led to an overproduction of unevenly distributed exopolysaccharides like in the Δsaci0446-saci1909 deletion mutant. Therefore, both Saci1908 and Saci1909 might be involved in exopolysaccharide production in S. acidocaldarius, but that only Saci1908 is in the Saci0446 regulatory network as the phenotype of Saci1909 deletion is independent of Saci0446.

In vitro DNA-binding assays of Saci0446

To test whether Saci0446 is directly involved in the regulation of the genes that were differentially expressed in the Δsaci0466 strain, Saci0446 was heterelogously expressed and purified (Supplementary Figure 6). Electrophoretic mobility shift assays were performed with similar sized (about 185 bp) DNA probes containing either promoter regions of potential target genes flaX (saci1177), flaB (saci1178), aapA (saci2314), saci1908 and the promoter region of the own gene (saci0446), or part of the saci0446-coding region as a specificity control (Figure 7a). Electrophoretic mobility shift assay with all probes resulted in the formation of multiple protein–DNA complexes with similar relative mobilities (Figure 7a). Binding affinities differed and were significantly higher for the promoter fragments (KD in the range of 100–600 nM) than for the fragment containing the coding sequence (Table 2). Except for the latter fragment, a positive binding cooperativity was observed with Hill coefficients exceeding 1 (Table 2). In particular, Saci0446 bound to both its own promoter region and the promoter of aapA with comparable and the highest affinities (KD=147 nM and 134 nM, respectively). Footprinting assays are described in the Supplementary Material (Supplementary Figure 7).

Figure 7
figure 7

In vitro DNA-binding analysis of Saci0446. Electrophoretic mobility shift assays of DNA binding of saci0446 to various DNA fragments. Lengths of the tested probes are 173 bp (p/o saci0446), 186 bp (ORF saci0446), 187 bp (p/o saci1177), 193 bp (p/o saci2314), 185 bp (p/o saci1178) and 181 bp (p/o saci1908). Protein concentrations are identical for all assays (in monomeric nM concentrations). The position of the free DNA probe is indicated as F.

Full size image
Table 2 Binding parameters of saci0446-DNA binding
Full size table

In conclusion, Saci0446 binds DNA with a high affinity but low sequence specificity and the affinity of binding to potentially targeted promoter regions is higher than for a non-relevant DNA sequence. Therefore, it is probable that Saci0446 mediates regulation by cooperative binding at the promoter loci.

Discussion

From studies using bacterial models, it is now well recognized that microbial biofilm development involves coordinated events leading to well-defined and distinct phenotypes that must obey a tightly regulated genetic program. Although archaea are frequently detected in biofilm communities, the molecular bases that underlie the sessile lifestyle remain to be discovered. In this study, we described the identification and the initial characterization of six homologous transcriptional regulators named as Lrs14 regulators from S. acidocaldarius regarding their potential role during biofilm development. Our analyses unraveled that three of them (saci1223, saci1242 and saci0446) are involved in regulating different aspects of biofilm development and that their mode of action presumably targets different pathways during this process in S. acidocaldarius. To our knowledge, this is the first report on regulatory components for the biofilm mode of growth in the archaeal domain.

In a previous study, we determined that the Lrs14 transcriptional regulator Saci1223 was one of the few common upregulated proteins among biofilm-grown Sulfolobus spp. (Koerdt et al., 2011), thus suggesting its role as a key regulatory factor in biofilm development. Indeed, the deletion of this regulator resulted in S. acidocaldarius cells that were noticeably impaired to build biofilms (Figures 2 and 3). As CLSM analysis showed Δsaci1223 was impaired in surface colonization (∼50% less cells on the surface), suggesting that the assembly of surface structures may be perturbed (Henche et al., 2011). However, none of the surface structure components, with the exception of upsAE, were found to be significantly altered in expression in biofilms formed by Δsaci1223 (Figure 4a). Although ups pili are essential for ultraviolet-induced cell aggregation and DNA transfer (Fröls et al. 2008), their role in surface attachment from shaking cultures and during biofilm maturation was demonstrated (Henche et al., 2011). However, Δups strains formed very unstable biofilms showing cell clusters unevenly distributed, which is different from Δsaci1223 biofilms. Therefore, the deletion of saci1223 might lead to a more complex pleiotropic response, which still has to be understood. In the future, additional analyses such as whole-transcriptional profiling of Δsaci1223 biofilms at different maturation stages could shed light into the role of Saci1223 during biofilm growth of S. acidocaldarius.

The deletion of saci0446 led to one of the most intriguing and distinct phenotypes. Δsaci0446 cells formed densely packed biofilms characterized by secretion of larger amounts of EPS as well as their impaired motility (Figures 3, 4 and 6). As the lack of archaella does not affect biofilm formation in S. acidocaldarius, its primary role is in motility and not persistence on surfaces (Henche et al., 2011). However, the upregulation of one pilin subunit (aapA), presumably resulting in increased numbers of the adhesive pili on the cell surface, could be the reason that the Δsaci0446 deletion strain showed a much denser appearance in the CLSM analysis (Figure 3). This taken together with the augmented EPS amount leads to the very stout and packed Δsaci0446 biofilms.

Saci0446 efficiently bound to the promoters of flaB and aapA genes and to its own promoter sequence, most likely acting as an activator for flaB and as repressor for aapA. In archaellum regulation, the FHA domain-containing protein ArnA and the vWA domain-containing protein ArnB interact strongly in vivo to consequently act as repressors of archaella expression (Reimann et al., 2012). The data presented here indicates Saci0446 as another player in archaellum regulation and implies that the archaellum transcriptional regulation network seems to be more complex than previously envisaged.

The lack of Saci0446 led to an increased production of EPS (Figure 5). This could be either due to increased exoploysaccharide production (Figure 3) and/or a significant change in the glycosylation pattern of cell surface-associated or extracellular proteins. As mentioned before, an exopolysaccharide biosynthetic pathway has not yet been described for any archaeon. Here, we identified a gene locus (saci1904–1927) whose gene products shared similarities with enzymes involved in bacterial exopolysaccharide biosynthesis and secretion (Figure 6a). Via genetic epistasis analysis we could determine that the deletion of saci1908 abolished EPS overproduction of the Δsaci0446 deletion strain (Figure 6c), thus strongly suggesting that its gene product has a relevant role in the EPS biosynthetic pathways of S. acidocaldarius. As Saci0446 bound to the promoter region of saci1908 and thereby would act as its repressor, as saci1908 mRNA levels were increased in the Δsaci0446 strain. Saci1908 is a putative membrane protein exhibiting 15 transmembrane regions with no function assigned so far. Therefore, it will be of great interest to unravel the function of Saci1908 regarding its role during EPS production/secretion.

The dual role of Saci0446 in controlling cell motility and formation of biofilms is reminiscent of several bacterial models in which tight regulation exists between cell motility and biofilm formation. For instance in Escherichia coli, CsgD is a key transcriptional regulator for curli production as well as a master regulator of biofilm formation. CsgD was shown to directly repress gene expression of flagella components, and activate the synthesis of extracellular polysaccharides, thereby switching from planktonic growth to the biofilm mode (Zogaj et al., 2001; Pesavento et al., 2008). Moreover, CsgD modulates adrA upregulation, which encodes one of the enzymes for cyclic di-GMP synthesis (Ogasawara et al., 2010). c-di-GMP consequently inhibits cell motility by interfering with the flagella motor speed via the c-di-GMP-binding protein YcgR (Wolfe and Visick, 2008; Boehm et al., 2010). From our results, we can suggest that Saci0446 acts in a rather opposite manner when compared with CsgD from E. coli, as Saci0446 activates archaella synthesis and represses genes required for biofilm formation.

In conclusion, this study reveals for the first time that members of the Lrs14 proteins, an archaea-specific class of transcriptional regulators, act in modulating the development of biofilms in S. acidocaldarius. In more detail, we demonstrated that saci0446 most likely represses EPS production during biofilm growth, while activating cell motility; thereby we propose to name saci0446 as abfR1 (for Archaeal Biofilm Regulator 1) as our results clearly show its role in building and shaping biofilm communities of S. acidocaldarius.