The Stl repressor from Staphylococcus aureus is an efficient inhibitor of the eukaryotic fruitfly dUTPase

DNA metabolism and repair is vital for the maintenance of genome integrity. Specific proteinaceous inhibitors of key factors in this process have high potential for deciphering pathways of DNA metabolism and repair. The dUTPase enzyme family is responsible for guarding against erroneous uracil incorporation into DNA. Here, we investigate whether the staphylococcal Stl repressor may interact with not only bacterial but also eukaryotic dUTPase. We provide experimental evidence for the formation of a strong complex between Stl and Drosophila melanogaster dUTPase. We also find that dUTPase activity is strongly diminished in this complex. Our results suggest that the dUTPase protein sequences involved in binding to Stl are at least partially conserved through evolution from bacteria to eukaryotes.

DNA metabolism and repair is vital for the maintenance of genome integrity. Specific proteinaceous inhibitors of key factors in this process have high potential for deciphering pathways of DNA metabolism and repair. The dUTPase enzyme family is responsible for guarding against erroneous uracil incorporation into DNA. Here, we investigate whether the staphylococcal Stl repressor may interact with not only bacterial but also eukaryotic dUTPase. We provide experimental evidence for the formation of a strong complex between Stl and Drosophila melanogaster dUTPase. We also find that dUTPase activity is strongly diminished in this complex. Our results suggest that the dUTPase protein sequences involved in binding to Stl are at least partially conserved through evolution from bacteria to eukaryotes.
DNA integrity and the fidelity of DNA replication are of vital importance. The dUTPase enzyme family, ubiquitous in free-living organisms [1] with some exceptions [2], contributes to these key issues by regulating the cellular dUTP/dTTP ratio. dUTPases catalyze the pyrophosphorolysis of dUTP into dUMP and pyrophosphate, providing the dUMP precursor for thymidylate biosynthesis. This enzymatic reaction also promotes clearance of dUTP from the cellular milieu, thereby preventing DNA polymerases from introducing dUMP moieties into DNA [1]. The significance of this sanitizing action is due to the fact that most DNA polymerases cannot distinguish between dUTP and dTTP and will readily utilize either of these two building blocks, depending only on their relative availability [3,4].
Elimination or inhibition of dUTPase activity leads to massive uracil incorporation into DNA that provokes futile hyperactivation of the base-excision repair pathway and results in DNA strand breaks followed by chromosome fragmentation and cell death [5,6]. This cell death pathway is usually referred to as 'thymine-less cell death' and may also be induced by chemotherapeutic drugs interfering with de novo thymidylate biosynthesis, such as fluoropyrimidines and methotrexate derivatives [7]. In fact, this chemotherapeutic strategy is frequently used clinically both against neoplastic diseases and against pathogenic microorganisms [3,[7][8][9][10]. Inhibition of dUTPase by small molecular drugs may also enhance the effectivity of this clinical protocol [11]. Several small molecular dUTPase inhibitors have been identified in the literature [12][13][14][15]. A proteinaceous dUTPase inhibitor, namely the staphylococcal Stl repressor, has also been discovered recently and it was shown to be active against trimeric dUTPases of several staphylococcal phages, as well as against the trimeric mycobacterial dUTPase [16][17][18].
Most probably, this interesting cross-species effect needs to be necessarily associated with structural features present in both phage and mycobacterial dUT-Pases. Notably, as most dUTPases belong to the all-b dUTPase enzyme family, the main structural fold is well preserved not just among prokaryotic dUTPases, but also in eukaryotic ones [19,20]. Within the evolutionary conserved dUTPase fold, three b-pleated polypeptide subunits form a trimeric enzyme possessing three equivalent active sites situated at the intersubunit clefts [21,22]. Although the overall conservation of the fold is clearly a major characteristic of the all-b dUTPase enzyme family, at the residue level only those residues are conserved that are directly involved in active site architecture [20,23]. Other protein surfaces potentially available for binding a macromolecular partner show great variation with respect to polarity, charge distribution, H-bonding, and Van der Waals capabilities. Therefore, it is an intriguing question to investigate whether any eukaryotic dUTPase may also form a protein-protein complex with the staphylococcal Stl. It is worthwhile to note that in the case of the enzyme family of uracil-DNA glycosylases, the UGI inhibitor protein (from the Bacillus subtilis phage PBS2) is fully functional in complexation and inhibition of not only prokaryotic, but also human and other eukaryotic uracil-DNA glycosylases, presenting a potentially relevant parallel situation [24][25][26].
In the dUTPase-Stl interaction investigated so far, functional effects of the complexation result not only in enzymatic inhibition of dUTPase, but also in perturbation of the repressor function of Stl [16,17,27]. In Staphylococcus aureus, Stl is responsible for repressing replication of SaPIbov1 pathogenicity island (SaPI) [28]. Staphylococcus aureus pathogenicity islands (SaPIs) are mobile genetic elements being responsible for horizontal gene transfer, a process being important for bacterial evolution [29,30]. Transcription of the SaPI may be induced upon helper phage infection by a specific interaction partner, which in the case of SaPIbov1 Stl is the helper phage dUTPase [27]. It was also shown that dUTPase removes Stl from its bound DNA [16][17][18].
In the present study, we wished to investigate whether Stl is able to form a stable complex with the eukaryotic Drosophila melanogaster dUTPase in vitro. Further on, we decided to determine the functional effects following from this complexation. Our results obtained by several independent methodologies show that a strong complex is formed between Stl and Drosophila dUTPase, similar to the case with phage and mycobacterial dUTPases. In this complex, dUTPase enzymatic activity is significantly reduced, but DNA binding to Stl may still be possible.

Protein expression and purification
The Stl-encoding gene sequence has been inserted into pGEX-4T-1 vector allowing glutathione S-transferase (GST) fusion expression and purification. The D. melanogaster dUTPase gene has been ligated into pET-15b vector between the BamHI and NdeI cleavage sites resulting in translation of a His-tagged dUTPase construct enabling purification with Ni-NTA affinity chromatography. Both constructs were expressed in Escherichia coli BL21 (DE3) Rosetta cells under similar conditions (cf. also [17,31,32]).
For protein expression, 0.5 L of LB medium was inoculated with a 5 mL overnight cell culture and grown at 37°C until OD 600 reached 0.5. At this point, protein expression was induced by the addition of 0.5 mM isopropyl b-D-1-thiogalactopyranoside. The cell cultures were incubated for further 4 h at 37°C for D. melanogaster dUTPase and 30°C for protein Stl. After centrifugation at 1376 g for 20 min at 4°C, cell pellets were resuspended in 15 mL precooled PBS and centrifuged again at 1376 g for 20 min at 4°C, then stored at À80°C until further usage.
Cells containing protein Stl in a GST-fused form were resuspended in 30 mL of 50 mM TRIS/HCl containing 1 M NaCl, EDTA-free protease inhibitor cocktail tablet, 10 mM dithiothreitol, 0.1 mgÁmL À1 DNase, and 0.01 mgÁmL À1 RNase A at pH 7.5. After sonication and centrifugation, the suspension was applied onto a glutathione column. Stl elution was carried out by cleavage of the column-bound GST tag using 80 units of thrombin in 3 mL reaction volume.

Size exclusion chromatography (SEC)
Size exclusion chromatography (SEC) was carried out on AKTA FPLC purification system using a Superose 12 10/ 300 GL column (GE Healthcare) previously equilibrated with dUTPase buffer (20 mM HEPES, 100 mM NaCl, 5 mM MgCl 2 , 10 mM b-mercaptoethanol, pH = 7.5). Complex formation was estimated based on comparison of the peak elution volumes of separate proteins with the value corresponding to the mixture of the two components. Fractions of 0.5 mL were collected after each injection. Peak elution fractions were concentrated on Amicon Ultra-4 ultrafiltration membranes (Merck-Millipore) before subsequent mass spectrometric analysis.

Differential scanning fluorimetry (DSF)
Samples were heated from 20 to 80°C using three parallels of each measurement. For visualization of protein unfolding, Sypro Orange protein dye was used in 1000-fold dilution. Melting points were obtained as global minimums corresponding to the first derivate of the melting curve.

Mass spectrometry
Mass spectra were measured in positive ion mode using a Waters QTOF Premier instrument with electrospray ionization source. Native conditions were applied; that is, ions were generated from aqueous 5 mM NH 4 HCO 3 buffer solution (pH: 7.8) containing the protein at 1 lM monomer concentration. Under such conditions, native protein complexes can be transferred from the solution to the gas phase. The capillary voltage was 2800 V, the sampling cone voltage was 128 V, and the temperature of the source was kept at 90°C. Mass spectra were recorded in the mass range of 1500-6000 m/z.

Enzyme activity and inhibition assay
These measurements were taken according to our previously used protocol [17]. Hydrolysis of dUTP results in proton release to the solution which can be followed as a change in pH. To quantify this, phenol red indicator was added to the reaction mixture (1 mM HEPES, 150 mM KCl, 5 mM MgCl 2 , 40 lM phenol red, pH: 7.5) and its absorbance was measured continuously as a function of time at 559 nm and 293 K using a 10-mm path length plastic cuvette. Initial velocity was determined from the first 10% of the progress curve. At least three parallel measurements were taken in all cases.
For Stl inhibition measurements, 100 nM of dUTPase and different amounts of protein Stl were preincubated together for 5 min in the measuring buffer at 293 K. The enzymatic reaction was always initiated by the addition of the dUTP molecule after mixing all other components.

Electrophoretic mobility shift assay (EMSA)
Electrophoretic mobility shift assay (EMSA) experiments were carried out on 8% TRIS/borate/EDTA (TBE) gels using a double-stranded 43-mer oligonucleotide (corresponding to the oligo termed 'Inter-R' in [33], with the sequence 'tcctcgaacaaattatctcacatcgagatatttatttcaacat' representing the Stl-specific DNA binding site. Samples were mixed in 'EMSA buffer' (TBE pH = 7.5, 100 mM NaCl, 0.5 mM EDTA). Before loading onto the gel, samples were incubated for 15 min at 293 K. After 1-h pre-electrophoresis of the empty gel on 150 V, the samples were run using the same voltage for 45 min at room temperature. GelRed was used to stain DNA. DNA bands were visualized by UVI-Tec gel documentation system after 15 min.

Native gel electrophoresis
Native gel electrophoresis was set up in a two-phase polyacrylamide gel. Acrylamide concentration was 4% in the stacking gel (pH = 6.8) and 10% in the resolving gel (pH = 8.8). After 30-min pre-electrophoresis without sample addition at constant 100 V on ice, the electrophoresis was performed for another 2.25 h at 200 V in native 'ELFO buffer' (30.3 gÁL À1 TRIS base, 144 gÁL À1 glycine, pH = 8.7). During electrophoresis, the whole apparatus was placed on ice. The gel was stained with Coomassie Brilliant Blue G250 dye (Thermo Fisher Scientific, Waltham, MA, USA).

Structural and homology modeling
Three-dimensional structural views were created by PYMOL (version 0.99rc6) [34]. The Clustal Omega server was used for multiple sequence alignment [35]. Drosophila melanogaster dUTPase structure was visualized as a homology model based on the human dUTPase crystal structure (PDB 3EHW) using the SWISS-MODEL server [36].

Results and Discussion
Staphylococcal Stl forms a stable complex with Drosophila dUTPase Figure 1 presents a structural alignment of one phage (S. aureus Φ11), one prokaryotic (Mycobacterium tuberculosis), and one eukaryotic (D. melanogaster) dUTPase [37,38]. It is clearly shown that the overall fold is well conserved both at the subunit and at the functional homotrimer level (Fig. 1A,B). However, surface representation shown at the same orientation for the three trimeric dUTPases (Fig. 1C,D,E) presents largely varied distribution of polar, charged, and hydrophobic surfaces. Despite this variation, the interaction among the mycobacterial dUTPase and Stl shows similar characteristics to the staphylococcal phage dUTPase-Stl interaction [16].
To study the potential binding of fruitfly dUTPase to Stl, first we applied size exclusion chromatography as a widely used straightforward technique to investigate protein-protein interactions. The chromatograms shown in Fig. 2A clearly indicated that a complex is formed in the mixture of the two protein components and this complex elutes at a position associated with higher molecular mass as compared to either of the other two components. The size exclusion chromatography experiment also allowed us to conclude that the complex of the two proteins is stable enough to be withheld in the complex state upon the dilution that necessarily occurs during the gel filtration process.
To check whether the presumed complex formation has any effect on protein stability, another well-characterized technique was used. Namely, we have performed differential scanning fluorimetry (DSF) experiments and found that indeed, the complexation induced a higher thermal stability, as the estimated melting temperature of 53.3°C for dUTPase and 57.3°C for Stl was shifted to 60.3°C in the mixture of the two proteins (Fig 2B).
While these two independent experiments provided unequivocal evidence for the physical contact between the two proteins (Stl from the prokaryote S. aureus and dUTPase from the eukaryote D. melanogaster), these data did not allow an estimation of the stoichiometry of the complex. To continue our investigations, we therefore initiated native gel experiments as in our earlier studies [39]. Data shown in Fig. 2C argue that on the one hand, the complex between staphylococcal Stl and D. melanogaster dUTPase is clearly visible on the native gel, while on the other hand, using different stoichiometric mixtures, bands corresponding to the separate components almost completely disappear at one exact component ratio being close to 3 : 2 dUTPase-Stl monomer assembly. Interestingly, this 3 : 2 stoichiometry was also suggested in the complex of the staphylococcal Φ11 phage dUTPase and Stl [17].
To further study whether this stoichiometry may be valid, we decided to analyze the gel-filtrated complex by mass spectrometry. Our aim was to compare the molecular ionic species present in the mixture of the two proteins to those molecular ions that are present in the separate solutions of the two proteins. Figure 3 upper panel shows the mass spectrum of the fruitfly dUTPase on its own. Mass data indicate that the trimeric fruitfly dUTPase can dissociate into monomers under the mass spectrometric conditions. This phenomenon was also observed in the case of a specific construct of Drosophila virilis dUTPase [40]. Mass spectra for the Stl protein on its own were already published [17] and showed the presence of both monomeric and dimeric Stl species under the experimental condition of native mass spectrometry. Mass spectra of the gel-filtrated complex of Stl and dUTPase-shown on the lower panel of Fig. 3-present unequivocal evidence for the complexation of these two proteins. Namely, a species with a molecular mass corresponding to the complex formed between one dUTPase trimer and one Stl dimer (or two Stl monomers) is observed. This stoichiometry is also in agreement with the data of the native gel experiment.

Functional effects of complex formation between Stl and Drosophila melanogaster dUTPase
Having found that fruitfly dUTPase and the staphylococcal Stl protein form a stable complex that resists experimental conditions of gel filtration, native gel electrophoresis, and mass spectrometry, we wished to determine the putative functional effects of this complexation. First, we tested whether Stl may inhibit the enzymatic activity of fruitfly dUTPase, as shown previously for phage and mycobacterial dUTPases.
As shown in Fig. 4A, there is a dose-dependent activity loss of dUTPase upon the addition of increasing concentrations of Stl to the reaction mixtures. In these experiments, the two proteins were preincubated before the addition of the dUTP substrate to start the enzymatic reaction, as previously in the case of phage and mycobacterial dUTPases [16,17].
The apparent IC 50 of Stl determined in these enzyme inhibition experiments was calculated to be 30 AE 5 nM. This value is in good agreement with the IC 50 of Stl determined for phage and mycobacterial dUTPase. However, the total inhibition observed at saturating Stl concentrations was~40%, to be compared with almost 100% inhibition for Φ11 phage dUTPase and 80% inhibition for the mycobacterial dUTPase [16,17]. These characteristics indicate that although the complexation proceeds similarly, the actual inhibitory capacity of Stl within the different complexes reflects some species-specific differences.
We also studied whether fruitfly dUTPase may perturb Stl-DNA complexation. In the EMSA experiment presented in Fig. 4B, we could nicely reproduce the previously published effect of Φ11 dUTPase on the DNA-Stl complex. Namely, the gel lanes show that increasing amount of Φ11 dUTPase leads to reappearance of the DNA band associated with the free DNA form (not bound to protein). However, when fruitfly dUTPase was added to the DNA-Stl complex, we could not observe dissociation of DNA from the DNA-Stl complex. Instead, a new band has appeared   at a higher position on the gel which we putatively identified as a ternary complex (DNA-Stl-dUTPase). Although the existence of this putative ternary complex needs further investigation, it is evident based on the presented data that the fruitfly dUTPase does not necessarily disrupt the interaction between Stl and its cognate DNA sequence.
Possible structural background of species-specific differences in Stl-induced dUTPase inhibition While all three dUTPases mentioned in this study can be inhibited by Stl in vitro, a remarkable difference is present in the maximal degree of their inhibition. To provide a putative explanation for these differences, we have aligned their sequences (Fig. 5). We have highlighted all amino acids possessing the same or at least similar side-chain characteristics-these segments may be accordingly involved in Stl binding (cf. also [18]). We were also looking for amino acid positions which are identical or similar in Φ11 and M. tuberculosis dUTPases but are of different characteristics in D. melanogaster dUTPase. Such residue alterations could serve as a basis for the observed significant difference in the degree of inhibition by Stl. The D. melanogaster dUTPase contains a Drosophila specific Cterminal extension, which may alter steric properties or surface charge characteristics of the enzyme. The Φ11 dUTPase has a phage-specific insert between the third and fourth sequence motifs [18], and the M. tuberculosis enzyme has a short mycobacteria-specific surface loop just before the fifth motif [41]. These factors may also be important for the differences in dUT-Pase interaction with protein Stl.

Conclusions
Protein-protein interactions can be conserved among species, especially if orthologue components of a given complex are present in the different species. In the present work, our focus was somewhat different: We investigated whether a eukaryotic representative of the evolutionary conserved dUTPase enzyme family may bind to a staphylococcal repressor (Stl) that is not present in other species. We found that D. melanogaster dUTPase and Stl form a strong complex with significant functional effect on dUT-Pase enzymatic activity, parallel to Stl-induced inhibition of phage and prokaryotic dUTPases. We conclude that Stl may be considered as a useful tool for specific inhibition of dUTPases in diverse systems. Further studies in progress in our laboratory will reveal whether in vivo dUTPase inhibition can be achieved in D. melanogaster model organism transfected with Stl.