StaR Is a Positive Regulator of Topoisomerase I Activity Involved in Supercoiling Maintenance in Streptococcus pneumoniae

The DNA topoisomerases gyrase and topoisomerase I as well as the nucleoid-associated protein HU maintain supercoiling levels in Streptococcus pneumoniae, a main human pathogen. Here, we characterized, for the first time, a topoisomerase I regulator protein (StaR). In the presence of sub-inhibitory novobiocin concentrations, which inhibit gyrase activity, higher doubling times were observed in a strain lacking staR, and in two strains in which StaR was over-expressed either under the control of the ZnSO4-inducible PZn promoter (strain ΔstaRPZnstaR) or of the maltose-inducible PMal promoter (strain ΔstaRpLS1ROMstaR). These results suggest that StaR has a direct role in novobiocin susceptibility and that the StaR level needs to be maintained within a narrow range. Treatment of ΔstaRPZnstaR with inhibitory novobiocin concentrations resulted in a change of the negative DNA supercoiling density (σ) in vivo, which was higher in the absence of StaR (σ = −0.049) than when StaR was overproduced (σ = −0.045). We have located this protein in the nucleoid by using super-resolution confocal microscopy. Through in vitro activity assays, we demonstrated that StaR stimulates TopoI relaxation activity, while it has no effect on gyrase activity. Interaction between TopoI and StaR was detected both in vitro and in vivo by co-immunoprecipitation. No alteration of the transcriptome was associated with StaR amount variation. The results suggest that StaR is a new streptococcal nucleoid-associated protein that activates topoisomerase I activity by direct protein-protein interaction.


Introduction
The Streptococcus pneumoniae chromosome is confined within the nucleoid, in which it is compacted by up to 1000-fold [1], allowing 1.3 Mb to be accommodated within cells just 1-2 µm long. Chromosome topology depends on DNA supercoiling (Sc), which is controlled by DNA topoisomerases [2]. S. pneumoniae possesses three of these enzymes: two type II topoisomerases (topoisomerase IV and gyrase), which cleave both DNA strands, and a single type I enzyme (topoisomerase I, TopoI), which cleaves only on one strand. Gyrase and TopoI actively regulate the degree of Sc solving topological problems associated with dynamic DNA remodeling in Escherichia coli [3][4][5][6] as well as in S. pneumoniae [7][8][9]. In addition, in other bacteria, a number of nucleoid-associated proteins (NAPs) [10] also control Sc, forming a functional network that maintains DNA topology via bending, wrapping, bridging and constraining supercoils [11]. Transcription also regulates Sc, given that

The Absence of StaR Affects Novobiocin Susceptibility in S. pneumoniae
To investigate the role of the hypothetical protein encoded by spr0929, herein named staR, we performed an InterPro analysis to search for bacterial homologues. The best hit found was the NA37 (Nucleotide-Associated proteins of 37 kDa) family. The YejK protein of E. coli, the representative member of this family and the only one with experimental information, had 19.4% identity and a 46.6% similarity with pneumococcal StaR ( Figure 1A). YejK has been described as a NAP [35]. To decipher the function of StaR, the growth of R6 (wild type) and the previously constructed ∆staR deletion mutant [34] was analyzed with or without NOV. NOV inhibits the ATPase activity of gyrase, which is necessary for its Sc activity. However, gyrase has relaxing activity in the absence of ATP hydrolysis, which could also contribute to the final effect in Sc relaxation induced by NOV [36,37]. While in the absence of NOV, both strains showed equivalent growth rates ( Figure 1B); in the presence of sub-inhibitory NOV concentrations (0.5 × MIC), ∆staR grew slower than R6: doubling time 112 ± 6 min versus 89 ± 6 min (p = 0.01), suggesting a role of StaR in Sc maintenance under challenging conditions. In order to purify and characterize StaR, a chromosomal fragment containing its coding gene (staR) was cloned into the E. coli pET28a BL21-CodonPlus (DE3) expression system, and overproduction of histidinetagged Star (H 6 -StaR) was achieved by induction with IPTG ( Figure 1C). A protein with an apparent molecular weight of 37.1 kDa, corresponding to the predicted molecular weight of the H 6 -StaR protein (39.3 kDa: 37 kDa + 6 His), was detected and further purified by affinity chromatography. Protein from fraction 10 in Figure 1C was collected and used in subsequent in vitro experiments.

The Expression Level of StaR Affects Novobiocin Susceptibility
To elucidate the role of StaR in NOV susceptibility, the ∆staRP Zn staR strain was constructed by introducing an ectopic copy of staR under the control of a promoter inducible by zinc (P Zn ) in the ∆staR strain. The growth of this strain was analyzed in the presence of 150 µM ZnSO 4 as the inductor of staR transcription and compared to R6 as control. In addition, the levels of StaR and topoisomerases involved in Sc homeostasis (TopoI and gyrase) were determined by Western blot along the growth curve. Quantification was performed using RpoB as an internal control since no change in its transcription was observed under NOV treatments [7]. While, in the absence of NOV, these strains showed equivalent growth rates both in the absence and in the presence of ZnSO 4 , in the presence of 0.5 × MIC of NOV, ∆staRP Zn staR grew slower than R6 (Figure 2A). The doubling times in the absence of ZnSO 4 were 215 min ± 32 and 94 min ± 10 for ∆staRP Zn staR and R6 (average ± SD, n = 3), respectively. In the presence of ZnSO 4 , these differences were similar: 154 min ± 19 for ∆staRP Zn staR and 89 min ± 10 for R6. Growth in the presence of 150 µM of ZnSO 4 yielded StaR increases of about 280-fold at all times considered, either in the presence or absence of NOV. However, StaR levels in strain ∆staRP Zn staR with ZnSO 4 were only about 2-fold higher than those in R6 ( Figure 2B,C). Thus, a 2-fold increase in the amount of StaR as well as its absence produced a higher susceptibility to NOV.
Regarding expression levels of topoisomerases, treatment with NOV decreased the level of TopoI about 4-fold at time points 90 and 150 min, and 6-fold at 210 min, in R6 compared to non-treated cells. This was expected considering previous results [7]. Similar decreases were observed in the strain ∆staRP Zn staR, which showed a reduction of about 7-fold in TopoI level in NOV-treated versus non-treated cells at all times tested (90, 150, and 210 min; Figure 2B,C). Therefore, the differences observed in NOV susceptibility between both strains were not associated with a topoisomerase unbalance, but they were due to the distinct levels of StaR as a consequence of its ectopic location. Therefore, appropriate levels of StaR are important for the Sc homeostatic response against the topological stress imposed by sub-inhibitory NOV concentrations [7].
To test whether higher levels of StaR affect cell viability, we used a plasmid cloning approach. We cloned staR into pLS1ROM, rendering pLS1ROMstaR, in which staR is under the control of the maltose-inducible P M promoter. Accordingly, StaR was induced in the presence of maltose independently on NOV treatment ( Figure 3B). The growth rate and protein levels of ∆staR strain containing pLS1ROMstaR were analyzed upon induction with maltose in the presence or in the absence of NOV. Growth in the presence of 0.4% sucrose + 0.4% maltose (SM) yielded a StaR increase of about 10-fold at 150 and 210 min with respect to sucrose (S)-grown cultures, although no effect on NOV susceptibility was observed ( Figure 3A). However, induction with 0.8% maltose (M) increased NOV susceptibility, a condition yielding the increase of StaR to about 30-fold (150 min and 210 min) with respect to S-grown cultures ( Figure 3B,C). StaR protein levels after induction with SM were similar to those observed in R6, consistent with the absence of any effect on growth ( Figure 3B). In contrast, treatment with M induced a 3.5-fold change at all times tested. Altogether, our data indicate that NOV susceptibility increased both in the absence of StaR (Figures 1 and 2A) and when its levels were increased by 2-to 3.5-fold (Figures 2 and 3).

The Expression Level of StaR Affects Novobiocin Susceptibility
To elucidate the role of StaR in NOV susceptibility, the ΔstaRPZnstaR strain was constructed by introducing an ectopic copy of staR under the control of a promoter inducible by zinc (PZn) in the ΔstaR strain. The growth of this strain was analyzed in the presence of 150 μM ZnSO4 as the inductor of staR transcription and compared to R6 as control. In coli YejK and S. pneumoniae StaR using multiple sequence alignment. Asterisks label identical residues between both proteins and colons denote similar residues. (B) Absence of staR affects NOV susceptibility. Isogenic strains carrying wild-type staR (R6) or a deletion of staR (∆staR) were grown to OD 620 nm = 0.3, diluted 200-fold in the absence or presence of NOV at 0.5 × MIC and their growth was recorded in a TECAN Infinite 200 PRO reader. Data are the average of three independent replicates ± SEM. (C) Purification of H 6 -StaR from E. coli BL21-CodonPlus (DE3) cells harboring pET28-staR. SDS-PAGE of 30 µL of crude extract samples of no-induced (CE) or IPTG-induced (CEI) cultures. Fraction numbers eluted from the AKTA prime system column are indicated; 30 µL of fractions were run in the gel that was stained with Coomassie-blue. Molecular weights are indicated on the left. and 210 min; Figure 2B,C). Therefore, the differences observed in NOV susceptibi tween both strains were not associated with a topoisomerase unbalance, but the due to the distinct levels of StaR as a consequence of its ectopic location. Therefo propriate levels of StaR are important for the Sc homeostatic response against th logical stress imposed by sub-inhibitory NOV concentrations [7].

The Supercoiling Level Is Affected by StaR under Novobiocin Treatment
As part of the Sc homeostasis response, a decrease in TopoI is produced after NOV treatment (Figure 3), coincidentally with a decrease in topA transcription [7]. The effect of StaR levels in DNA Sc was analyzed by measuring Sc density of pLS1 in ΔstaRPZnstaR with or without NOV. We analyzed the topoisomer distribution of the internal plasmid pLS1 by two-dimensional agarose electrophoresis, a suitable approach for studying Sc levels, which is a reflection of nucleoid compaction [7]. This technique allows separation of DNA molecules by mass and shape as an estimation of the chromosomal Sc. As expected, treatment of ΔstaRPZnstaR with NOV resulted in DNA relaxation. In the absence These results suggest that StaR has a direct role in resolving DNA relaxation stress and that its levels need to be precisely regulated. In fact, staR is located in a DOWN domain in the R6 genome, and its transcription decreases more than 2-fold under NOV treatment [7]. This explains the increased NOV susceptibility observed when staR is ectopically located in ∆staRP Zn staR into a locus (bgaA) not located in a supercoiling domain, as well as when it is overproduced in the presence of ZnSO 4 or maltose.

The Supercoiling Level Is Affected by StaR under Novobiocin Treatment
As part of the Sc homeostasis response, a decrease in TopoI is produced after NOV treatment (Figure 3), coincidentally with a decrease in topA transcription [7]. The effect of StaR levels in DNA Sc was analyzed by measuring Sc density of pLS1 in ∆staRP Zn staR with or without NOV. We analyzed the topoisomer distribution of the internal plasmid pLS1 by two-dimensional agarose electrophoresis, a suitable approach for studying Sc levels, which is a reflection of nucleoid compaction [7]. This technique allows separation of DNA molecules by mass and shape as an estimation of the chromosomal Sc. As expected, treatment of ∆staRP Zn staR with NOV resulted in DNA relaxation. In the absence of StaR expression (i.e., no ZnSO 4 in the medium), the value of negative DNA Sc density (σ) was −0.056, and it decreased to −0.051 and −0.049 at 0.5 × MIC and 1 × MIC NOV, respectively ( Figure 4A). However, under StaR overexpression (150 µM ZnSO4) this variation in σ was higher: from −0.057 to −0.048 and −0.045 at 0.5 × MIC and 1 × MIC NOV, respectively ( Figure 4B). Therefore, under inhibitory concentrations of NOV, negative Sc density was higher without StaR (σ = −0.049) than with a 2-fold overproduction of StaR (σ = −0.045), p = 0.01. These results suggest that StaR increases the level of NOV-induced DNA relaxation.

StaR Is a Nucleoid-Associated Protein
To investigate the possible association of StaR with the nucleoid, we performed Stimulated Emission Depletion (STED) super-resolution microscopy of the wild-type strain (R6). After fixation, the nucleoids were stained with Sytox orange and StaR was immunofluorescently labeled. A total of 66.2% of cells (n = 4912) expressed detectable levels of StaR. In these, co-localization was readily observed, indicating that StaR localizes to the

StaR Is a Nucleoid-Associated Protein
To investigate the possible association of StaR with the nucleoid, we performed Stimulated Emission Depletion (STED) super-resolution microscopy of the wild-type strain (R6). After fixation, the nucleoids were stained with Sytox orange and StaR was immunofluorescently labeled. A total of 66.2% of cells (n = 4912) expressed detectable levels of StaR. In these, co-localization was readily observed, indicating that StaR localizes to the nucleoid ( Figure 5A). To confirm co-localization, images of these cells were used to calculate the Pearson's correlation index ( Figure 5C). Thresholds corresponding to the average intensity of the green signal of the ∆staRP Zn staR strain (0.042) and a Pearson correlation index of 0.5 were applied. Under these criteria, StaR co-localized with the nucleoid in 51.5% (1674) of the R6 cells analyzed ( Figure 5C). As a control, ∆staRP Zn staR cells were imaged under the same growth, fixation and staining conditions in the absence of ZnSO 4 to image StaRdepleted cells ( Figure 5B). Consistently, the distribution of the background green signal detected drastically differed compared to that of R6. The ∆staRP Zn staR strain showed brilliant green areas without any conserved pattern that could be attributed to nonspecific binding of antibodies, and a low-intensity green signal which could correspond to cellular auto-fluorescence ( Figure 5B). Moreover, under the same co-localization criteria that was applied to the wild-type (R6) strain, none of the ∆staRP Zn staR cells analyzed showed any co-localization ( Figure 5C). This further reinforces the co-localization analysis and shows that, like its homolog YejK, StaR associates with the nucleoid and must be considered a new NAP in S. pneumoniae. nucleoid ( Figure 5A). To confirm co-localization, images of these cells were used to calculate the Pearson's correlation index ( Figure 5C). Thresholds corresponding to the average intensity of the green signal of the ΔstaRPZnstaR strain (0.042) and a Pearson correlation index of 0.5 were applied. Under these criteria, StaR co-localized with the nucleoid in 51.5% (1674) of the R6 cells analyzed ( Figure 5C). As a control, ΔstaRPZnstaR cells were imaged under the same growth, fixation and staining conditions in the absence of ZnSO4 to image StaR-depleted cells ( Figure 5B). Consistently, the distribution of the background green signal detected drastically differed compared to that of R6. The ΔstaRPZnstaR strain showed brilliant green areas without any conserved pattern that could be attributed to nonspecific binding of antibodies, and a low-intensity green signal which could correspond to cellular auto-fluorescence ( Figure 5B). Moreover, under the same co-localization criteria that was applied to the wild-type (R6) strain, none of the ΔstaRPZnstaR cells analyzed showed any co-localization ( Figure 5C). This further reinforces the co-localization analysis and shows that, like its homolog YejK, StaR associates with the nucleoid and must be considered a new NAP in S. pneumoniae.

StaR Specifically Activates S. pneumoniae TopoI
In vivo results of Sc variation in the presence or absence of StaR suggest a role of this protein in TopoI activation, since DNA relaxation in the presence of NOV is higher when StaR is overproduced. Alternatively, this protein could further inhibit DNA gyrase lead- . y-axis represents the mean intensity of green label (StaR) for ∆staRPZnstaR (red points) and R6 (blue points) and the x-axis indicated the Pearson correlation between green and red (nucleoids) labels. Gray lines in the xand y-axis correspond to the thresholds applied to calculate the number of cells where StaR co-localizes with the nucleoid. A yellow square highlights the R6 population showing StaR-nucleoid co-localization.

StaR Specifically Activates S. pneumoniae TopoI
In vivo results of Sc variation in the presence or absence of StaR suggest a role of this protein in TopoI activation, since DNA relaxation in the presence of NOV is higher when StaR is overproduced. Alternatively, this protein could further inhibit DNA gyrase leading to the same observations.
The effect of StaR on the enzymatic relaxing activity of TopoI was tested. Results showed that StaR activates the relaxation of pBR322 by TopoI in vitro, depending on the amount of StaR used ( Figure 6A). We have quantified the activation of TopoI by measuring both decrease of CCC plasmid form and increase of topoisomers ( Figure 6B). Since the gel running conditions do not allow for differentiation between OC and RC, both topoisomers with intermediate linking numbers as well as OC/RC forms were considered for calculating TopoI activity. The results showed the simultaneous decrease of CCC and increase of topoisomers as the StaR concentration increases. The CCC and topoisomers reached 2.5% and 97.5%, respectively, at the highest StaR concentration. Moreover, activation of TopoI by StaR depended on the incubation time ( Figure 6C). In contrast, StaR did not show topoisomerase activity by itself at concentrations equal to or higher than those used in TopoI activity assays ( Figure 6D). In contrast, no activation of S. pneumoniae DNA gyrase activity was observed, since 73% of CCC plus topoisomers was observed at all StaR concentrations tested (from 0.25 to 4 µM). Although an excess of StaR over TopoI is required for activation, these results suggest that StaR is a specific activator of TopoI.

Physical Interaction between StaR and TopoI
The specific activation of TopoI by StaR suggested a protein-protein interaction between them. To verify this, in vitro co-immunoprecipitation assays (Co-IP) were carried out with the purified proteins. The two proteins were incubated together and immuneprecipitated using polyclonal antibodies against either TopoI or StaR. The presence of TopoI in the anti-StaR pull-down fraction and the presence of StaR in the anti-TopoI pull-down fraction ( Figure 7A) indicated a direct interaction between these proteins in vitro.
We also performed Co-IP in vivo after formaldehyde crosslinking ( Figure 7B). We used extracts from cultures of ∆staRP Zn staR grown in either the absence or presence of ZnSO 4 , i.e., in the absence or presence of StaR. Western blots showed that StaR-Ab was able to pull down TopoI. Three bands were observed in the Western blot when the StaR-antibody was used in the Co-IP: a band presumably corresponding to TopoI (79 kDa), the band corresponding to StaR (37 kDa) and another band with an apparent molecular weight of 116 kDa, which may correspond to the TopoI-StaR complex. The last band corresponded to 26% of the total amount of the immunoprecipitated StaR. However, no pulldown of StaR by TopoI-Ab was observed, suggesting that the polyclonal antibody against TopoI dissociates the TopoI-StaR complex.

No Global Changes in Transcriptome Are Induced by the Absence or Presence of StaR
The transcriptome of the strain ∆staRP Zn staR was analyzed either in the absence of StaR (no ZnSO 4 added) or in its presence (150 µM ZnSO 4 ). In the presence of the P Zn inductor, 34 differentially expressed genes (DEGs) were detected (Table 1), with the same number of genes (17) up-or down-regulated. As expected, genes known to be regulated by high concentration of ZnSO 4 [7] were found among these DEGs, with seven up-regulated and six down-regulated. Up-regulated genes included the psaBCA operon coding a Mn/Zn transporter, czcD coding a cation efflux protein, prtA coding a virulence factor, and adhB involved in carbon metabolism, while celA was down-regulated. Consistent with the increased SatR levels detected in Western blots (280-fold increase, Figure 2) upon ZnSO 4 induction, mRNA levels of staR also increased about 400-fold. vation of TopoI by StaR depended on the incubation time ( Figure 6C). In contrast, S did not show topoisomerase activity by itself at concentrations equal to or higher th those used in TopoI activity assays ( Figure 6D). In contrast, no activation of S. pneumon DNA gyrase activity was observed, since 73% of CCC plus topoisomers was observed all StaR concentrations tested (from 0.25 to 4 μM). Although an excess of StaR over To is required for activation, these results suggest that StaR is a specific activator of Topo  weight of 116 kDa, which may correspond to the TopoI-StaR complex. The l responded to 26% of the total amount of the immunoprecipitated StaR. How down of StaR by TopoI-Ab was observed, suggesting that the polyclonal anti TopoI dissociates the TopoI-StaR complex.  When the effect of NOV on the transcriptome in the absence of StaR (−Zn) was compared with its effect when StaR was overexpressed (+Zn), no important differences were observed ( Figure 8A). A total of 324 (110 up-regulated) and 313 (108 up-regulated) DEGs were detected in the absence or presence of StaR, respectively, with 214 DEGs common to both conditions ( Figure 8A). However, a slightly differential response was observed, since about one-third of the DEGs under each condition was specific. In the absence of StaR, NOV triggered the expression of 110 genes (67 down-regulated and 43 up-regulated) that did not change in the presence of StaR. Similarly, when StaR was over-expressed, NOV changed the expression of 99 genes (58 down-regulated and 41 upregulated) that did not change when staR was repressed. A comparison of the functional classes of the DEGs did not present significant differences between both conditions. In addition, no difference in the location of DEGs along the genome was detected in the absence or presence of ZnSO 4 ( Figure 8B). the location of DEGs along the genome was detected in the absence or presence of ZnSO4 ( Figure 8B). These results suggest that the overexpression of StaR does not affect global transcription induced by NOV in a chromosomal-domain manner, and thus discard StaR as a transcriptional regulator.   1 to 2,038,615). A fold change ≥ 2 (absolute value) and a p-value-adjusted ≤0.01 were considered.
These results suggest that the overexpression of StaR does not affect global transcription induced by NOV in a chromosomal-domain manner, and thus discard StaR as a transcriptional regulator.

Discussion
The existence of protein networks governing DNA chromosomal topology of pathogens remains an unknown field. Sc levels, nucleoid compaction and viability in S. pneumoniae are determined by the balance between gyrase and TopoI activities [7]. This bacterium triggers homeostatic transcriptional responses when Sc density decreases by 25% [8] or when it increases by 40% [16]. These global responses affect transcription of topoisomerases. Relaxation triggers up-regulation (about 2-fold) of gyrase genes (gyrA and gyrB) and down-regulation of both TopoI (topA, around 10-fold) and TopoIV (parE and parC, around 3-fold) [7]. When Sc increases, topA is again down-regulated (about 2-fold), while gyrA, gyrB, parE and parC remain unchanged [16]. In this way, the down-regulation of topA transcription, although to different levels, allowed cell growth and the recovery of Sc.
In addition, the transcriptomic response of wild-type strain R6 to DNA relaxation, which is triggered by NOV, also involved down-regulation (about 2-fold) of the gene coding StaR (spr0929), which is indeed included in a down-regulated domain [8]. A similar decrease in transcription (around 3-fold) was observed for the gene coding HU, the only NAP described so far in S. pneumoniae, which is essential for Sc maintenance [32]. However, the down-regulation of HU and StaR only occurred at inhibitory NOV concentrations. No change in the transcription or amount of StaR (Figure 2) was observed at sub-inhibitory NOV concentrations. Nevertheless, the location of StaR in a down-regulated domain is important for cell survival in the presence of NOV (Figure 1). This location is consistent with the homeostatic response to relaxation, which involves the down-regulation of topA [7], and to the decrease of TopoI levels detected in Western blots ( Figure 2). Therefore, relaxation triggers down-regulation of both topA and staR. This would contribute to the recovery of supercoiling levels after relaxation, since the levels of StaR, which would normally activate TopoI, would decrease. Accordingly, the overproduction of StaR observed in ∆staRP Zn staR upon addition of ZnSO 4 reduces recovery of Sc at inhibitory (1 × MIC) concentrations of NOV: 21% of Sc variation after 30 min of NOV treatment versus 12.5% without induction (Figure 4). This is presumably due to TopoI activation by StaR. The role of StaR would be the activation of TopoI by protein-protein interaction. Although this activation would hypothetically affect transcription, a global effect was not detected in the in the RNA-Seq experiments carried out in the present paper at the concentration of NOV used.
S. pneumoniae lacks most of the NAPs characterized in E. coli, with the exception of HU [32] and SMC [33]. Using DAPI to stain DNA and immunostaining for NAPs, it has been shown in E. coli that HU and other NAPs localize to the nucleoid [38]. We have recently used super-resolution confocal microscopy and staining of the bacterial nucleoid with the DNA intercalant Sytox [9]. This technique allowed us to accurately identify and locate nucleoids and determine their level of compaction [9]. In this study, we took advantage of this technique to demonstrate that StaR indeed localizes in the nucleoid ( Figure 5). A statistical analysis performed with more than 2500 cells showed co-localization of StaR and the nucleoid in more than 50% of analyzed bacteria. Although we cannot compare this number with previous reports, given that, to our knowledge, no statistical quantification has been performed previously for any NAP, we consider that a co-localization higher than 50% indeed reflects a true co-localization. It is important to notice that cultures were not synchronous, and the localization of StaR in the nucleoid could depend on the cell cycle phase. This is the first demonstration of a NAP localizing to the nucleoid in S. pneumoniae.
Therefore, StaR should be considered a new NAP, which regulates the activity of DNA topoisomerases. Other NAPs regulating topoisomerases have been identified in other bacteria. This is the case of GyrI, which inhibits E. coli gyrase [39], GapR from Caulobacter crescentus that stimulates gyrase and Topo IV to remove (+) Sc during DNA replication [40], and HU that activates Topo I in M. tuberculosis [41]. E. coli YejK, homologous to StaR, has been shown to inhibit both gyrase and TopoIV activities [42]. We have not found gyrase activity alteration by StaR, which probably reflects the low sequence identity between YejK and StaR (Figure 1). Activation of Topo IV has not been tested, given that S. pneumoniae Sc is mostly controlled by the opposing activities of Topo I and gyrase [7,16]. Regulation of the topoisomerase type I activity has also been observed in eukaryotes. It has been described that TOP1 activity is controlled by the phosphorylated RNAPII to modulate elongation during transcription [43]. In S. pneumoniae, TopoI is the main topoisomerase involved in the regulation of supercoiling [7,16] and in transcription [14]; therefore, its activity must be finely regulated.
While co-membership to the same Pfam family does not imply sharing exactly the same properties, we have confirmed that StaR is a NAP and has a similar size to E. coli YejK, indicating that NA37 is a proper global name for this protein family. In future studies, the structure-function relationship may add support at residue-level to the observations described here. Unfortunately, no homologous structural templates for modeling are currently available. Numerous attempts to resolve the StaR structure from S. pneumoniae and Enterococcus faecalis were unsuccessful since crystals showed low-resolution diffraction. In all, evidence indicate NA37 proteins are not prone to experimental structural analyses. In contrast, novel ab initio modeling tools based on artificial intelligence such as AlphaFold achieve unprecedented accuracy [44]. Such homology-free models may help to evaluate critical traits of StaR. These include the co-location of basic residues able to interact with DNA, the domain delineation, and the identification of interacting sides for docking studies.
To summarize, StaR is a new NAP of S. pneumoniae and is the first characterized TopoI modulator in this bacterium, playing an important role in the genome-relaxation-triggered homeostasis. Therefore, we propose that StaR is involved in the fine-tuning control of the activity of the essential TopoI, contributing to the homeostatic response to Sc changes.

Bacterial Strains, Growth Conditions and Transformation
S. pneumoniae strains were grown as static cultures at 37 • C in a casein hydrolysatebased liquid medium (AGCH) containing 0.2% yeast extract and 0.3% sucrose [45] (C + Y). E. coli was grown in LB medium at 37 • C with agitation. A strain with a deletion in staR was constructed as described [34]. Transformation of pneumococcal strains was performed as previously reported [45] and selection of transformants was in 1 µg/mL tetracycline, 250 µg/mL kanamycin, 2.5 µg/mL chloramphenicol, or 1 µg/mL erythromycin. Growth was followed by measuring the OD at 620 nm either with a UV-visible spectrophotometer (Evolution 201, Thermo Scientific, Waltham, MA, USA) or in a microplate reader (Infinite F200, Tecan, Männedorf, Switzerland). Both measures correlate linearly by means of the equation y = 0.2163x + 0.1151 (y = microplate reader measure, x = spectrophotometer), R 2 = 0.98 [32].

Construction of Strains and Plasmids
To overproduce and purify StaR, a PCR fragment containing its coding gene was amplified from R6 genomic DNA with primers spr0929_F (5 -GCGCGCTAGCATGGATATTTATA TTAAGAAAGCC) and spr0929_R (5 -GCGCGGATCCTTATTTACTTTGGATATCCTCG) [34]. Restriction enzyme sites are showed underlined in all oligonucleotides. The amplicon was digested with NheI and BamHI and cloned into pET28a cut with the same enzymes. The plasmid was transferred to BL21-CodonPlus (DE3) (Agilent Technologies, Santa Clara, CA, USA) E. coli cells. To test the physiological effects of the overexpression of staR, a genetic fusion between the P Zn -ribosome binding site and the staR coding region was made to render the ∆staRP Zn staR strain. The staR gene was amplified with spr0929_ScIF (5 -: GCGCGAGCTCATGGATATTTATATTAAGAAAGCC) and spr0929_SlIR (5 -GCGCGTCGACTTATTTACTTTGGATATCCTCG) oligonucleotides using R6 chromosomal DNA as template. The amplicon was digested with SacI and SalI enzymes, targets included in the oligonucleotides, and introduced into pZK plasmid cleaved with the same enzymes to render plasmid pZK_0929. The pZK plasmid contains the upstream region of the czcD gene that shows up-regulation in the presence of zinc [46]. A PCR fragment containing the Zn-inducible promoter upstream the spr0929 ORF was amplified from pZK_0929 with KmR_B1_Rv (5 -CGCGGGATCCAGGATCCATCGATACAAATTCC) and pZK_Xb1_Rv (5 -GCGCTCTAGACACCATAAAAAATGAACTTGG) oligonucleotides. This construction was introduced into the spr1806 locus of the ∆staR strain. The spr1806 gene codes for a surface protein, which in laboratory conditions, is not expressed; subsequently, the deletion mutant shows no relevant differences with respect to the wild-type strain [47]. For that, DNA fragments of~1 Kb corresponding to the exact flanking regions of spr1806 were amplified with oligonucleotides spr1806_ExF (5 -CATAGGTCTCCACATCGAAA) and spr1806_LiURB1 (5 -GCGCGGATCCAACAGCTAGAAAATTCTATTCT), for the upstream region, and oligonucleotides spr1806_LiDFXb (5 -GCGCTCTAGATACTGAACCATCTGGGAATG) and spr1806_ExR (5 -TTGCCCGTCAAAATGATTTG) for the downstream region. Amplicons were cut with BamHI and XbaI, respectively, and simultaneously ligated to the fragment containing the Zn-inducible promoter and spr0929 ORF. The ligation fragment was introduced into pre-competent ∆staR cells by natural transformation and transformants were selected with 50 µg/mL kanamycin. Recombination was checked by PCR.
To test the role of staR overproduction, the gene was cloned into pLS1ROM vector under the control of P Mal promoter, which is controlled by MalR transcriptional repressor that is cloned in pLS1ROM under the control of the constitutive P tet promoter [48]. For this purpose, a PCR fragment containing staR was obtained from R6 chromosome using oligonucleotides 0929HindF (5 -CGCAAGCTTGGAGGACTTTTATGGATATTTATATTAAGAAAGC) and 0929BamR (5 -GCGGATCCGCAAAAAGGAAGACTGCTAGTACAAG), cut with HindIII and BamHI, and ligated to pLS1ROM cut with the same enzymes. The constructed plasmid pLS1ROMstaR was transformed into ∆staR strain and transformants were selected with 1 µg/mL erythromycin.

Cloning, Expression and Purification of Proteins
H 6 -StaR was purified by His-affinity chromatography using HiTrap Chelating HP columns (Amersham-Merck, Madrid, Spain) and an AKTA prime system (Amersham Bioscience, Amersham, UK) from E. coli BL21-CodonPlus (DE3) cells harboring pET28-StaR. The strain was grown in LB medium containing 50 µg/mL kanamycin at 37 • C to OD 600 nm = 0.6 and staR transcription was induced with 1 mM isopropyl thio-β-D-galactoside (IPTG) for 1 h. The cells were harvested by centrifugation, washed with phosphate-buffered saline (PBS) and suspended in 72 mL of buffer A (50 mM Tris-HCl pH 8.0, 0.5 M NaCl). The cells were sonicated (8 cycles of 20 sec ON/60 sec OFF) and cell extracts were clarified by centrifugation at 100,000× g for 1 h at 4 • C. A two-step protein fractionation with ammonium sulfate at 40% pellet and 60% saturation was performed. The pellet from 60% saturation was suspended in 51 mL of buffer A containing 40 mM imidazole and applied to a 5 mL HiTrap column equilibrated buffer A with 40 mM imidazole. Protein elution was achieved with 100 mL of a 100-to-500-mM imidazol gradient in buffer A, and fractions were analyzed by SDS-PAGE. Imidazol was removed from samples by dialysis against 20 mM Tris-HCl pH 7.5, 1 mM EDTA, 0.1 M NaCl, 50% glycerol. Protein concentration was determined with a Qubit 4 fluorimeter using the Qubit ® Protein Assay Kit (Invitrogen-ThermoFisher Scientific, Waltham, MA, USA). TopoI, GyrA and GyrB were purified by affinity chromatography in a Ni-NTA (Quiagen, Hilden, Germany) column following manufacturer's instructions as described previously [26].

Confocal Microscopy
The cells were grown until OD 620 nm = 0.4, centrifuged and suspended in PBS to a number of 8 × 10 9 cells/mL. The cells were fixed with 2% formaldehyde for at least 16 h. For immunofluorescence microscopy, 8 × 10 7 cells were incubated with 5 µM SytoxTM Orange Nucleic Acid Stain (Invitrogen) at room temperature for 5 min and extended into a poly-L-lysine coated glass slide. The cells were permeabilized by immersing the slide in methanol at −20 • C for 10 min. The slide was first incubated with 2% BSA, 0.2% Triton X100 in PBS (BSA-PBST), incubated with a rabbit StaR polyclonal antibody at 1/100 dilution in BSA-PBST buffer, washed with PBS and incubated with the anti-rabbit Abberior ® STAR 488 antibody (Abberior, Göttingen, Germany) at 1/200 dilution. After a wash with PBS, the slides were mounted with ProLongTM Gold Antifade Mountant (Invitrogen) and sealed. Nucleoids and StaR were imaged as previously described [9] in a confocal microscope STELLARIS 8-FALCON/STED (Leica Microsystems, Wetzlar, Germany) with a HC PL APO 100×/1.40 NA × OIL immersion objective. Super resolution images were acquired by Stimulated Emission Depletion (STED) microscopy using 660 nm depletion laser. Analysis was performed with Cell profiler software v4.2.5, and co-localization-correlation index was calculated for each bacteria with a custom-made routine.

Western Blot Assays and Antibody Purification
Whole cell lysates (~5 × 10 5 cells) were obtained by centrifugation of 10 mL cultures (OD 620 nm = 0.4). They were suspended in 400 µL of sample loading buffer (0.3 M Tris-HCl pH 6.8, 10% SDS, 50% v/v glycerol and 0.05% bromophenol) supplemented with 0.5 M β-mercaptoethanol and incubated for 5 min at 100 • C. Lysates were separated on Any kD™ Criterion™ TGX Stain-Free™ Protein Gels (Bio-Rad, Hercules, CA, USA). They were transferred to 0.2 µm PVDF membranes with a Trans-Blot Turbo Transfer System (Bio-Rad) at 25 V, 1 A for 30 min. Membranes were blocked with 5% skim milk in Trisbuffered saline for 2 h and incubated with anti-TopoI and anti-GyrA [16], anti-RpoB [14] and anti-StaR (diluted 1:1000). Rabbit polyclonal antibody against StaR was obtained from Davids Biotechnologie from 0.5 mg of protein extracted from SDS-gel following a 28-day SuperFast immunization protocol. Anti-rabbit IgG-peroxidase Ab (Sigma-Aldrich-Merck, Madrid, Spain) was used as the secondary Ab. SuperSignal West Pico chemiluminescent substrate (Thermo-Fisher, Waltham, MA, USA) was used to develop the membranes. Signal was detected with a ChemiDocTM MP system (Bio-Rad). Images were analyzed using Image LabTM software (Bio-Rad). Molecular masses of GyrA and TopoI are 92 kDa and 79 kDa, respectively.

DNA Topoisomerase Assays
Relaxation reactions of pBR322 by TopoI were carried out exactly as described previously [49]. Reactions of 200 µL contained 0.5 µg of CCC pBR322 in 20 mM Tris-HCl pH 8, 100 mM KCl, 10 mM MgCl 2 , 1 mM DTT, 50 µg BSA/mL and TopoI at the indicated concentrations. Incubation with TopoI was at 37 • C for 1 h and the reaction was terminated by 2 min incubation at 37 • C with 50 mM EDTA. Then an additional incubation of 1 h at 37 • C with 1% SDS, 100 µg/mL proteinase K was performed. Reaction products were ethanol precipitated, suspended in electrophoresis loading buffer and analyzed by electrophoresis in 1% agarose gels run at 18 V for 18 h. DNA quantification of agarose gels was done by scanning densitometry after electrophoresis and ethidium bromide staining. Quantification of TopoI activity was calculated by gel densitometry using the Image Lab program (Bio-Rad Laboratories, Hercules, CA, USA). To calculate activity, the OC form amount was determined and divided by the total amount of DNA in each well. Enzymatic reactions with gyrase were conducted as described previously [50].

Co-Immunoprecipitation (Co-IP) Assays
For in vitro Co-IP, the buffer used along the experiment was Co-IP buffer (10 mM sodium phosphate pH 7.4, 50 mM NaCl, 0.5% Nonidet P40). Purified TopoI (60 ng/µL) was dialyzed against the buffer, and StaR (9.2 µg/mL) was diluted 1000-fold in the same buffer. A total of 2 µg of TopoI and StaR, respectively, was mixed into 200 µL of buffer for 2 h at 4 • C with rotation at room temperature. Purified Abs (4 µg) were bound to Dynabeads ® Protein G for 2 h in the buffer at room temperature with rotation as described by the manufacturer and were added to the protein mix. After incubation, the mix was washed 3 times and it was collected from the magnetic grid with SDS-loading buffer without β-mercaptoethanol. The samples were boiled for 5 min, loaded in a 4-20% SDS-polyacrylamide gel and run for 1 h at 170 v. Gels were stained with Coomassie blue for about 1 h.
For in vivo Co-IP, 10 mL of fixing buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl, 0.5 mM EGTA, 1 mM EDTA, 11% (v/v) formaldehyde) was added to 100 mL of culture grown to OD 620 nm = 0.4 (7 × 10 9 cells) with or without 150 µM ZnSO 4 . This mix was incubated for 30 min at room temperature. Crosslinking was stopped by adding 10 mL of cold quenching solution (1.25 M glycine in 50 mM Tris-HCl pH 8.0, 100 mM NaCl, 0.5 mM EGTA, 1 mM EDTA). The mix was transferred to an ice/water bath and rotated for 30 min at 4 • C. Cells were collected by centrifugation at 4 • C and suspended in 50 mL ice-cold PBS.
TopoI and StaR antibodies were purified as described previously [14]. The antibodies (10 µg) were bound to 50 µL Dynabeads ® Protein G. For immunoprecipitation, 500 µL of sonicated suspension (antigens) was added to 200 µL of the magnetic-bead-Ab complex and incubated on a rotating mixer for 4 h at 4 • C. The tube was placed on a magnet and the supernatant was removed. It was washed once with the lysis buffer, once with the lysis buffer containing 500 mM NaCl and once with the lysis buffer containing 250 mM LiCl. The supernatant was removed and proteins bound to Dynabeads were resupended in 30 µL of the sample loading buffer containing 50 mM glycine. Dynabeads ® Protein G-antigen was incubated at 65 • C overnight with shaking to elute DNA from the beads and reverse crosslinks. Samples were fractionated in SDS/PAGE and proteins were detected by Western blot using TopoI-Ab (1:500 dilution) and StaR-Ab (1:1000 dilution) antibodies.

Analysis of the Topology of Plasmids
Plasmid DNA topoisomers were analyzed in neutral/neutral two-dimensional agarose gels. The first dimension was run at 1.5 V/cm in a 0.4% agarose (Seakem-Lonza, Basel, Switzerland) gel in 1 × Tris-borate-EDTA (TBE) buffer for 20 h at room temperature. The second dimension was run at 7.5 V/cm in 1% agarose gel in 1 × TBE buffer for 7-9 h at 4 • C. Chloroquine (Sigma-Aldrich-Merck, Madrid, Spain) was added to the TBE buffer in both the agarose and the running buffer. Chloroquine is a DNA intercalating agent that removes negative Sc in bacterial plasmids. Increasing the concentration of chloroquine progressively eliminates negative Sc until the plasmid is relaxed and can then introduce net positive Sc. In this way, the use of adequate concentrations of chloroquine during each dimension in the 2D analysis allows the efficient resolution of the different topoisomers [51]. Chloroquine concentrations used were 1 µg/mL and 2 µg/mL in the first and second dimension, respectively. Gels were stained with 0.5 µg/mL ethidium bromide for 1 h at room temperature. Images were captured in a ChemiDoc Imaging System (Bio-Rad) and analyzed with the Image Lab software (BioRad). DNA supercoiling density (σ) was calculated from the equation σ = ∆Lk/Lk 0 . Linking number differences (∆Lk) were determined using the equation ∆Lk = Lk − Lk 0 , in which Lk 0 = N/10.5, being N the size of the molecule in bp (4408) and 10.5 the number of bp per one complete turn in B-DNA. To simplify it, σ = Lk of the most abundant topoisomer/(N/10.5). The Lk of the most abundant topoisomer was calculated taking into account that the topoisomer that migrated with ∆Lk = 0 in the second dimension has a ∆Wr = −14 during the first dimension (the number of positive supercoils introduced by 2 µg/mL chloroquine).

RNA Library Preparation for RNA-Seq
Total RNA (5 µg) obtained from samples containing 2 to 4 × 10 8 cells with the RNeasy mini kit (Quiagen) was depleted of ribosomal RNA using the Ribo-Zero Magnetic Kit (Bacteria) (Illumina, San Diego, CA, USA). Libraries for RNA sequencing (RNA-Seq) were prepared using the ScriptSeq v2 RNA-Seq system (Illumina). Briefly, 1 µg of RNA samples were chemically fragmented. Using randomly primed cDNA synthesis, cDNAs that were tagged at the 5 end (equivalent to the 3 end of the original RNA) were synthesized. The cDNAs were then tagged at the 3 end using Terminal-Tagging Oligos (TTO). These oligonucleotides randomly annealed to the cDNA and were extended by DNA polymerase. The resulting di-tagged cDNAs were purified with the Ampure bead XP system. The enrichment and barcoding of the purified di-tagged cDNAs were done with 15 cycles of PCR. The library size was determined with a 2100 Bioanalyzer Instrument (Agilent Technologies). qPCR quantifications were done with a Kapa Library quantification Kit (Kapa Biosystems, Sigma-Aldrich-Merck, Madrid, Spain).

RNA Seq Data Analysis
Analysis of RNA-Seq data was carried out using the web-based platform Galaxy [14]. The quality of raw sequence data was analyzed with the FASQC tool. Sequencing reads were mapped against the S. pneumoniae R6 genome (ASM704v1) using the BWA software package (Galaxy version 0.7.17.4) in simple Illumina mode. The number of reads overlapping each coding gene was obtained using program feature Count. Count tables were used as input in DESeq2 for the analysis of differential expression. A threshold p-value-adjusted of 0.01 was considered.