Molecular analysis of the UV-inducible pili operon from Sulfolobus acidocaldarius

Upon ultraviolet (UV) stress, hyperthermophilic Sulfolobus species show a highly induced transcription of a gene cluster responsible for pili biogenesis: the UV-inducible pili operon (ups operon). This operon is involved in UV-induced pili assembly, cellular aggregation, and subsequent DNA exchange between cells. As the system increases the fitness of Sulfolobus cells after UV light exposure, we assume that transfer of DNA takes place in order to repair UV-induced DNA damages via homologous recombination. Here, we studied all genes present in the ups cluster via gene deletion analysis with a focus on UpsX, a protein that shows no identifiable functional domains. UspX does not seem to be structurally essential for UV-induced pili formation and cellular aggregation, but appears to be important for efficient DNA transfer. In addition, we could show that pilin subunits UpsA and UpsB probably both function as major pilin subunits in the ups pili.


Introduction
Upon ultraviolet (UV) stress Sulfolobus species show a high upregulation of a gene cluster encoding proteins responsible for formation of type IV pili (T4P) (Fröls et al. 2007;Götz et al. 2007): the ups operon (UV-inducible pili operon of Sulfolobus) (Fröls et al. 2008). Indeed, when analyzed by transmission electron microscopy (TEM), a large number of pili could be observed on the surface of UV-stressed Sulfolobus cells from different species (Ajon et al. 2011). An additional phenotypic characteristic of UV-stressed Sulfolobus cells is the formation of large cellular aggregates (Fröls et al. 2008), which was shown to be mediated by ups pili in a species-specific manner (Ajon et al. 2011). Moreover, Sulfolobus cells can exchange chromosomal DNA in a pili-dependent manner which was shown to increase cellular fitness under UV stress (Ajon et al. 2011). Transfer of DNA therefore probably plays a role in repair of double-strand breaks (DSBs) caused by UV radiation as was proposed in Ajon et al. (2011). Because the aggregation was also shown to be inducible by the DNA strand-break-inducing agent bleomycin, the first trigger for pili formation and subsequent aggregation is thought to be the sensing of DSBs in the DNA (Fröls et al. 2008). The mechanism behind this process remains unknown. DNA exchange mechanisms among hyperthermophiles for repair of DNA have been described in more detail by van Wolferen et al. (2013).
In other organisms, T4P were also shown to mediate DNA transfer (Filloux 2010). Different studies showed that the pili facilitate the uptake of extracellular DNA in competent Gram-positive and Gram-negative bacteria. The exact role of T4P in competence is still not well understood, but their presence is essential for successful DNA uptake (reviewed in Kr€ uger and Stingl 2011). It has been suggested that DNA is brought close to the cell surface by binding to the pili that subsequently retract. A recent study for the first time reported the binding of a pilin subunit (minor pilin ComP) of competent Neisseria meningitidis to specific sequences of self-DNA suggesting that pili indeed bring DNA to the cell surface (Cehovin et al. 2013). Besides having a role in competence, two examples are known in which T4P are involved in a conjugative system. First, the IncI1 conjugative plasmid R64 in Escherichia coli that carries the pil genes (Kim and Komano 1997;Yoshida et al. 1999;Komano et al. 2000), and second, PAPI-1 DNA which conjugates between Pseudomonas species and also comprises pil genes. The latter is located on a pathogenicity island that has been obtained via horizontal gene transfer (Carter et al. 2010). A role of T4P in cellular chromosomal exchange and subsequent repair of damaged DNA has never been shown before. Moreover, unlike the described bacterial T4P, the archaeal ups genes are not present on a conjugative plasmid or acquired via horizontal gene transfer.
The UV-induced pili system was initially studied in S. solfataricus (Fröls et al. 2008;Ajon et al. 2011), but because of the availability of genetic tools for manipulating the Sulfolobus acidocaldarius genome (Wagner et al. 2009(Wagner et al. , 2012 we switched to the latter organism. There is a high conservation and functional similarity among all sequenced Sulfolobus species. The ups operon encodes five proteins: UpsX, a hypothetical protein with no conserved regions; UpsE, an ATPase; UpsF an integral membrane protein; and UpsA and UpsB, two putative pilin subunits containing class III signal peptides.
Here, we have analyzed the different roles of the genes in the ups cluster. The secretion ATPase UpsE and the membrane protein UpsF are, as expected, essential for the formation of ups pili. Interestingly, UpsX is not essential for pili formation, but the deletion of its gene resulted in decreased DNA exchange suggesting a role in DNA transfer. Pilin subunits UpsA and B are essential for cellular aggregation. However, as deletion mutants of either of the subunits still formed pili, it appears that a mixed structure is essential for cellular aggregation.

Culture conditions
Sulfolobus acidocaldarius strains MW001, MR31, and JDS183 and derived mutants were grown aerobically at 78°C in basic Brock medium (Brock et al. 1972), supplemented with 0.1% NZ amine AS (Sigma, Munich, Germany), 0.2% dextrin, and 20 lg/mL uracil, and adjusted to pH 3.5 with sulfuric acid. For solid media the medium was supplemented with 1.5% gelrite. Plates were incubated for 5-6 days at 78°C. E. coli-competent cells DH5a and ER1821 (NEB, Frankfurt am Main, Germany), used for, respectively, cloning and methylation of plasmid DNA, were grown in Lysogeny broth medium (10 g/L tryptone; 5 g/L yeast extract; 10 g/L NaCl) at 37°C supplemented with the appropriate antibiotics. Growth of cells was monitored by optical density measurements at 600 nm.

Ultraviolet treatment, aggregation assays
Ultraviolet light treatment was performed as described in Fröls et al. (2008). Ten mL culture (OD 600 0.2-0.3) was treated with a UV dose of 75 J/m 2 (254 nm, UV crosslinker; Spectroline, Westbury, NY) in a plastic petri dish. Subsequently cultures were incubated at 78°C for 3 h. Samples taken at different time points were analyzed with phase contrast microscopy, survival rate assays, and electron microscopy. To quantify aggregated cells after induction with UV, 5 lL of cell culture (diluted to OD 0.2) was spotted on a microscope slide covered with a thin layer of 1% agarose in Brock minimal medium. A coverslip was added when the drop had dried. Cells were visualized with phase contrast microscopy. Free and aggregated cells (≥3) were counted for at least three fields per strain using ImageJ cell counter (NIH, Bethesda, MD). Percentages of cells in aggregates were subsequently calculated.

Deleting/tagging genes in S. acidocaldarius
To construct deletion and gene replacement strains, upand downstream flanking areas of the genomic regions of interest (~600 bp) were amplified with primers listed in Table S1. Primers were designed according to the genomic sequence of Sulfolobus acidocaldarius DSM639. Overlap polymerase chain reaction (PCR) was performed to connect the up-and downstream fragments (Wagner et al. 2012). The PCR product was subsequently cloned into pSVA406, carrying an ampicillin resistance gene, which resulted in the plasmids summarized in Table 1. The plasmids were methylated in E. coli ER1821-containing pM.EsaBC4I (NEB) (Kurosawa and Grogan 2005) and transformed into S. acidocaldarius MW001/MR31 (Wagner et al. 2012). Integrants were selected on plates lacking uracil and grown in 24-well plates for 2 days in the same medium. Subsequently cultures were plated and grown for 5 days on second selection plates containing uracil and 100 lg/mL 5-fluoroorotic acid to select for clones in which the plasmid looped out by homologous recombination. Obtained colonies were tested by PCR for successful deletion/replacement of the genes. Correctness of strains was confirmed by DNA sequencing. Strains that were made during this study are listed in Table 2.
qPCR and operon mapping on S. acidocaldarius cDNA To compare expression of the ups operon from different strains (UV/not-UV induced), RNA was isolated from 10 mL cultures (from MW001 and mutants, Table 2) using TriFast TM (Peqlab, Erlangen, Germany). DNA was subsequently degraded by incubating the RNA with DNA-seI (RNAse free, Fermentas, St. Leon-Rot, Germany) according to the manufacturer's protocol. Proper DNA degradation was confirmed by performing a PCR with primer pair 2033 + 2087 (Table S1) on the RNA. cDNA synthesis was performed on 1 lg of RNA with the First Strand cDNA Synthesis Kit (Fermentas). Random primers were used and the manufacturer's protocol was followed. Quantitative PCR (qPCR) was performed using the Maxima SYBR Green/ROX qPCR master mix. qPCR primers were designed for saci1493-saci1496b (upsX-upsB), they have a melting temperature around 60°C and give a product of 80-150 bp in length (primers 2073-2082, Table  S1). As a control, primers for an lrs14 gene that was not found to be differentially expressed after induction with UV in microarray studies were used. Control qPCR was performed according to the manufacturer's instructions. The obtained CT values were used to compare non-UVinduced with UV-induced expression of the tested genes. Moreover, expression was compared between MW001 and deletion strains. Differences in expression were displayed as log 2 folds.
To determine whether the genes in the ups gene cluster are present as an operon, isolated cDNA of UV-induced S. acidocaldarius MW001 was used to amplify intergenic regions between upsX and E; E and F; F and A; and A and B using primer pairs: 2073 + 2076; 2020 + 2013; 2077 + 2080; and 2038 + 2082, respectively (Table S1).

DNA transfer assays
DNA transfer between S. acidocaldarius cells was assayed by selecting prototrophic (pyr + ) recombinants of two pyrE mutant strains. The ups + pyrE strains were MR31 (wt1) and JDS183 (wt2). Strain MR31 has an 18-bp deletion in pyrE (nt 154-171). Strain JDS183 contains a frame shift A. Gosh and S.V. Albers, unpubl. data mutation (duplication of T) at nt 44. The DupsE strains were SA1 (DupsE1) and DG253 (DupsE2). SA1 was derived from background strain MR31, and DG253 contains a transition mutation (A-G) at pyrE nt 335. The DupsX strain was GA09, which contains a frameshift mutation in pyrE (Dbp 28, see Table 2). Liquid cultures were grown at 78°C and harvested at OD 600 0.15-0.35. Pellets were resuspended to a cell density of about 2 9 108 cells/mL. UV irradiation was performed as described before (Ajon et al. 2011). Recombination was assayed by spreading a mixture with 50 lL of each of two suspensions on selective plates without uracil. Plates were incubated for 5-6 days at 78°C as was described previously (Ajon et al. 2011). For each experiment the results were normalized, taking mixture wt1(UV) 9 wt2 (C) as 100%.

Electron microscopy analysis
Ultraviolet-induced pili in S. acidocaldarius cells and derived mutants were observed with TEM. Specimens were negatively stained with 2% uranyl acetate on carbon-coated copper grids. Microscopy was performed with a Philips CM10 electron microscopy operated at 120 kV. Images were recorded using a Gatan 4K CCD camera at different magnifications.

Results
Bioinformatics and transcriptional analysis on the ups operon As described before (Fröls et al. 2008), the ups operon encodes five proteins that are together thought to build the type IV pilus (Fig. 1): UpsE, a secretion ATPase; UpsF, an integral membrane protein; UpsA and UpsB, two putative pilin subunits containing class III signal peptides; and UpsX, a protein with unknown function. UpsX is a predicted cytoplasmic protein, but in contrast to UpsE, F, A, and B, no specific domains could be predicted by BLAST (Altschul et al. 1990), SMART (Schultz et al. 1998), and HHpred (Söding 2005) (data not shown). An alignment of UpsX sequences from different Sulfolobales revealed several conserved regions and amino acids (Fig. S1). As the bioinformatics did not predict a possible UpsX function, it was subjected to further functional analysis. Synteny analysis (SyntTax; Oberto 2013) revealed that all sequenced Sulfolobales contain a ups operon (Fig. 1), including recently sequenced Stygiolobus azoricus (A. Wagner and S.-V. Albers, unpubl. data). However, as Acidianus hospitalis, member of the Sulfolobales, lacks genes encoding pilin subunits UpsA and UpsB, this species probably does not build functional ups pili. In addition, the membrane protein encoding upsF seems to be incomplete in this species. It is therefore likely that A. hospitalis lost part of its ups operon throughout evolution. Deep sequencing on S. acidocaldarius cDNA (O. Wurtzel, unpubl. data) revealed a transcriptional start site (TSS) in front of upsX (with~10,000 transcript reads), and TSSs in front of the upsE (160 reads) and upsA (133 reads) genes. These results suggest a primary TSS in front of upsX and secondary TSSs in front of upsE and upsA. Operon mapping using RT-PCR indeed shows that transcripts of upsX-B are connected and therefore probably present as one long transcript (Fig. 2). This is confirmed by the observation that a deletion of both the upsX gene and its promoter (ΔupsX -40) leads to an unpiliated phenotype, whereas a clean deletion of upsX does not (Fig. 6). Interestingly, the deep sequencing data revealed several antisense TSSs in upsB, which might have a regulatory function. Importantly, these deep sequencing data show already transcription numbers without prior UV induction of the samples, indicating that the operon is expressed highly even without UV stress. The ups system therefore probably shows a high basal activity. Previous microarray studies showed an upregulation of the transcription of all ups genes with log 2 folds of up to 3 (Fröls et al. 2007;Gö€ otz et al. 2007). By means of qPCR, this induction could be confirmed showing even higher log 2 folds of between 4 and 5 (Fig. 3).

Deletion mutant analysis of genes in the ups operon
In order to obtain more insights in the individual roles of the genes in the ups operon, markerless deletion mutants   Table S1. were created as described previously (Wagner et al. 2012) ( Table 2). The genotypes were confirmed by PCR and sequence analysis (data not shown). In addition, transcription levels of ups genes in the different deletion mutants were compared with those from wild-type cells by qPCR (Fig. 4). Generally, no strong polar effects could be observed on downstream genes of the deleted genes of interest. However, the deletion of upsE resulted in a lower transcription of upsF. Moreover, in a upsF mutant transcription of upsA and upsB was slightly reduced. The latter might possibly be due to the partial removal of the possible promoter region of upsA by the deletion of upsF. Growth curves and microscopy revealed wild-type growth and a normal cellular phenotype for all deletion mutants (data not shown). In deletion mutants of upsE, F, A, and B, no UV-induced aggregation could be observed (Fig. 5). This was to be expected as the secretion ATPase (UpsE), membrane protein (UpsF), and pilin subunits (UpsA/B) are all thought to be essential for pili formation. In line with this, no ups pili formation could be observed for the upsE and F deletion strains (Ajon et al. 2011;Fig. 6, and data not shown). As it is unknown if and how the two predicted pilin subunits UpsA and UpsB together build up one filament, we also looked at single-deletion mutants of upsA and upsB. To not confuse ups filaments with other surface structures (Henche et al. 2012;Jarrell and Albers 2012), these strains were made in an archaella-and aap (archaeal adhesive) pili-less background (MW501). When the single mutants of upsA and upsB were analyzed with EM, both strains showed pili formation, although at smaller numbers. These results suggest that both subunits are capable to form filaments, but that for efficient pili formation both genes are needed, likely resulting in mixed subunit pili.  Interestingly, the ΔupsX strain shows wild-type cellular aggregation upon UV treatment and also the ups pili are wild type like (Figs. 5 and 6). UpsX therefore does not seem to have an essential role in the pili formation or cellular recognition (aggregation).

Chromosomal marker exchange
To study a putative role of UpsX in DNA exchange between Sulfolobus cells, DNA transfer assays were performed. Series of auxotrophic ups/pyrE double mutants were used in mating experiments (Table 2). Two parental strains were mixed together and upon exchange of chromosomal DNA, pyrE mutations could be restored via homologous recombination resulting in prototrophic colonies as described previously (Ajon et al. 2011). For each experiment the results were normalized, taking the mix-ture of wt1 (UV) 9 wt2 (C) as 100% (Fig. 7). A mixture of two ups wild-type strains resulted in the formation of recombinants (Fig. 7, wt1 9 wt2, green bar), indicating DNA exchange between the two strains and confirming a high basal activity of the system, even without UV stress. Moreover, upon induction of one of the two strains with UV light, exchange of DNA increased four-to fivefold (Fig. 7, wt1 9 wt2, red bars). A upsE deletion strain did not support transfer of DNA when treated with UV light (Fig. 7, wt1 9 ΔupsE2, first red bar). Only when the wild-type strain (wt1) was induced with UV in this mixture, a significant increase in DNA exchange could be observed, showing that only one of the two strains needs to assemble pili for DNA exchange to occur. A mixture of two ΔupsE strains resulted in no DNA transfer (ΔupsE1 9 ΔupsE2). These results confirmed the previously observed essential function of ups pili for DNA  exchange between Sulfolobus cells (Ajon et al. 2011). However, when mixing the wt2 with UV-induced ΔupsX, still a higher amount of recombinants were formed than without UV induction, although this increase was only about 50% of that of a wt1 9 wt2 mixture. Similarly, a mixture of ΔupsX with ΔupsE resulted in a significantly lower increase in DNA exchange upon UV induction of the ΔupsX strain. These results imply that UpsX plays a role in the process of UV-induced DNA transfer between Sulfolobus cells, but that it is not essential. Possibly, UpsX plays a direct or indirect role in DNA transfer or processing of incoming/outgoing DNA.

In vivo protein levels and localization of UpsX
To localize UpsX in the cells, S. acidocaldarius upsX was genomically tagged with a C-terminal HA tag (MW101, Table 2). This strain was induced with UV and samples were taken 0, 1, 2, and 3 h after UV irradiation. Cytosol (C) and membranes (M) were subsequently separated using ultracentrifugation. Western blotting analysis on the different fractions confirmed that UpsX-HA migrates with the expected size of around 79 kDa on SDS-PAGE, with some smaller degradation products present in the cytosol fractions (Fig. 8). The 55 kDa protein was confirmed to be unspecific as it was also present in the negative control (À, lysate of MW001). Therefore, it was used as an internal loading standard. After UV induction, a clear increase in UpsX protein in line with the transcriptional response could be observed (Fig. 3). Interestingly, without UV induction, UpsX was found only in the cytosol while after UV stress, a significant fraction of about 35% of the protein seemed to localize to the membrane (Fig. 8).

Discussion
T4P have shown to be involved in numerous functions both in bacteria as well as in archaea. In Sulfolobales, different type IV pili have been described, their functions include: motility, attachment to surfaces, biofilm formation, and DNA exchange. In S. acidocaldarius, three different T4P involved in these functions can be found: archaeal adhesive (aap) pili (Henche et al. 2012), archaella (Jarrell and Albers 2012), and ups pili (Fröls et al. 2008;Ajon et al. 2011). Previous studies have shown the involvement of ups pili in UV-induced cellular aggregation and DNA exchange. This mechanism is unique to the Sulfolobales and is proposed to be involved in repair of DNA DSBs (Fröls et al. 2008;Ajon et al. 2011). Here, we have studied the individual genes of the S. acidocaldarius ups gene cluster in more detail using bioinformatics, transcriptional analyses, deletion mutant analyses, and localization experiments.
Synteny analysis revealed that all Sulfolobales, but no other species contain a ups cluster encoding: UpsX, a protein with unknown function; UpsE, a secretion ATPase; UpsF, a membrane protein; and UpsA and B, two class III signal peptide-containing pilin subunits. Expression and induction of this operon has been shown for a number of Sulfolobus species (Ajon et al. 2011). The ups system of A. hospitalis is almost certainly not active as it lacks both pilin subunits and part of the upsF gene, suggesting that it lost these genes during evolution. A possible explanation of this loss might be that Acidianus species do not encounter as much UV light as other Sulfolobales and therefore have slower mutation rates. For instance, A. hospitalis is a facultative anaerobe (unlike obligate aerobic Sulfolobus and Metallosphaera species), may thus grow in deeper, darker, areas of hot springs (Giaveno et al. 2013) being less exposed to UV light. In other Sulfolobales, the ups genes can be readily deleted with no apparent effect on vegetative growth. However, the presence of ups pili does have a clear beneficial effect on the vitality of Sulfolbus cells upon severe DNA damaging conditions (Ajon et al. 2011). The ups system might therefore have evolved to conquer extremely DNA damaging conditions such as high temperatures, low pH values, and high UV doses. The increased exchange of DNA would make regular homologous recombination more efficient by increasing the chances of having a homologous template.
Operon mapping confirmed that the ups cluster of S. acidocaldarius is transcribed as one transcript demonstrating that it is truly an operon. Indeed, a deletion of the primary promoter region in front of upsX aborted pili formation, whereas a clean deletion of upsX has no effect. Deep sequencing data, moreover, showed the presence of additional TSSs within the operon in front of upsE and upsA (O. Wurtzel, unpubl. data). Secondary promoter elements were also found in the S. acidocaldarius archaella operon (Lassak et al. 2012) and might be essential to fine tune the stoichiometry of the different proteins within the ups pili: for example, many more pilin subunits will be needed than membrane proteins involved in the assembly of a functional pilus. The strong upregulation of the ups operon upon UV light was confirmed with qRT-PCR.
Deletion mutants of genes in the ups operon revealed that upsE and upsF are individually essential for pili formation as well as aggregation further expanding on the observation with the upsEF double-deletion mutant (Ajon et al. 2011). Deletion mutants of the pilin subunit genes upsA and upsB no longer showed UV-induced aggregation. Interestingly though, both strains still formed ups pili, but in lower numbers. These results suggest that UpsA and UpsB are both major pilin subunits that in wild-type cells might form a mixed pilus structure, this in contrast to bacterial T4P systems which were so far found to have only one major pilin subunit (Ayers et al. 2010;Giltner et al. 2012). In bacteria, minor and major pilins have been implicated in adherence to various different surfaces, cells from the same species, and host tissues (reviewed by Giltner et al. 2012). Also ups subunits are thought to be involved in strain-specific recognition and interaction between cells from the same species (Ajon et al. 2011). Preliminary data suggest that a specific region in the pilin subunit UpsA is responsible for the recognition of glycosylated S-layer proteins from cells from the same species (M. van Wolferen and S.-V. Albers, unpubl. data).
UpsX is a protein with so far unknown function, but it is highly conserved in all species that contain a ups operon suggesting that it fulfills an essential role in UVdependent DNA transfer. Intriguingly, however, a deletion mutant of upsX still formed ups pili, and the UV-induced cellular aggregation was comparable to wild type. UpsX therefore does not seem to play a structural role in the ups pilus nor is it essential for cellular recognition or assembly of the pili. Possibly other functions of UpsX would be DNA transfer, and indeed, a UV-induced upsX deletion mutant contributed less to DNA transfer compared to the wild-type strain. These data indicate that UpsX promotes DNA exchange but it is not essential for this process. Interestingly, a significantly larger portion of the UpsX protein localized to the membrane fraction in time upon UV exposure. Possibly, UpsX promotes DNA exchange at the membrane via a yet unknown mechanism. Future biochemical experiments might provide more insights into the actual function of UpsX.