Discovery of the role of a SLOG superfamily biological conflict systems associated protein IodA (YpsA) in oxidative stress protection and cell division inhibition in Gram-positive bacteria

Bacteria adapt to different environments by regulating cell division and several conditions that modulate cell division have been documented. Understanding how bacteria transduce environmental signals to control cell division is critical to comprehend the global network of cell division regulation. In this article we describe a role for Bacillus subtilis YpsA, an uncharacterized protein of the SLOG superfamily of nucleotide and ligand-binding proteins, in cell division. We observed that YpsA provides protection against oxidative stress as cells lacking ypsA show increased susceptibility to hydrogen peroxide treatment. We found that increased expression of ypsA leads to cell division inhibition due to defective assembly of FtsZ, the tubulin-like essential protein that marks the sites of cell division. We showed that cell division inhibition by YpsA is linked to glucose availability. We generated YpsA mutants that are no longer able to inhibit cell division. Finally, we show that the role of YpsA is possibly conserved in Firmicutes, as overproduction of YpsA in Staphylococcus aureus also impairs cell division. Therefore, we propose ypsA to be renamed as iodA for inhibitor of division. IMPORTANCE Although key players of cell division in bacteria have been largely characterized, the factors that regulate these division proteins are still being discovered and evidence for the presence of yet-to-be discovered factors has been accumulating. How bacteria sense the availability of nutrients and how that information is used to regulate cell division positively or negatively is less well-understood even though some examples exist in the literature. We discovered that a protein of hitherto unknown function belonging to the SLOG superfamily of nucleotide/ligand-binding proteins, YpsA, influences cell division in Bacillus subtilis by integrating metabolic status such as the availability of glucose. We showed that YpsA is important for oxidative stress response in B. subtilis. Furthermore, we provide evidence that cell division inhibition function of YpsA is also conserved in another Firmicute Staphylococcus aureus. This first report on the role of YpsA (IodA) brings us a step closer in understanding the complete tool set that bacteria have at their disposal to regulate cell division precisely to adapt to varying environmental conditions.

ligand-binding proteins, in cell division. We observed that YpsA provides protection 23 against oxidative stress as cells lacking ypsA show increased susceptibility to hydrogen 24 peroxide treatment. We found that increased expression of ypsA leads to cell division 25 inhibition due to defective assembly of FtsZ, the tubulin-like essential protein that marks 26 the sites of cell division. We showed that cell division inhibition by YpsA is linked to 27 glucose availability. We generated YpsA mutants that are no longer able to inhibit cell 28 division. Finally, we show that the role of YpsA is possibly conserved in Firmicutes, as 29 overproduction of YpsA in Staphylococcus aureus also impairs cell division. Therefore, 30 we propose ypsA to be renamed as iodA for inhibitor of division. 31 32 IMPORTANCE 33 Although key players of cell division in bacteria have been largely characterized, the 34 factors that regulate these division proteins are still being discovered and evidence for 35 the presence of yet-to-be discovered factors has been accumulating. How bacteria 36 sense the availability of nutrients and how that information is used to regulate cell 37 division positively or negatively is less well-understood even though some examples 38 exist in the literature. We discovered that a protein of hitherto unknown function 39 belonging to the SLOG superfamily of nucleotide/ligand-binding proteins, YpsA, 40 influences cell division in Bacillus subtilis by integrating metabolic status such as the 41 availability of glucose. We showed that YpsA is important for oxidative stress response 42 in B. subtilis. Furthermore, we provide evidence that cell division inhibition function of 43 YpsA is also conserved in another Firmicute Staphylococcus aureus. This first report on 44 the role of YpsA (IodA) brings us a step closer in understanding the complete tool set 45 that bacteria have at their disposal to regulate cell division precisely to adapt to varying 46 environmental conditions. Quantification of H2O2-treated cells revealed that 27% of WT and 79% of ΔypsA cells 108 were sick (n =100). To test if this phenotype is specifically due to absence of YpsA, we 109 introduced an inducible copy of ypsA at an ectopic locus. In the presence of inducer, fluorescence of YpsA-GFP showed that YpsA assembles into discrete foci (Fig. 3D). 128 Time-lapse microscopy conducted at 2 min interval for 10 min revealed that YpsA foci 129 are highly dynamic (Fig. 3I-L). Since YpsA-GFP retains fluorescence as a focus and that 130 the foci are mobile, and focus disruption occurs in some YpsA mutants (Fig. 6), we 131 conclude that the foci are not artifacts of non-functional misfolded aggregates. 132 a similar sporulation assay and found that cells overexpressing ypsA (98%) or ypsA-gfp 159 (127%) also displayed sporulation frequency similar to WT. To fully comprehend how 160 filamentous cells achieve WT-like sporulation efficiency, we observed the cell 161 were not filamentous, YpsA foci still formed in CH (Fig. 5F). 168 169 We hypothesized that lack of nutrients in CH compared to LB might be the reason for 170 lack of filamentation. To test our hypothesis, we externally added 1% glucose to the CH 171 medium. Intriguingly, cells grown in CH in the presence of glucose and inducer to 172 overproduce YpsA or YpsA-GFP lead to filamentation [YpsA: 3.31 ± 0.79 µm (Fig. 5C) 173 vs 6.49 ± 2.95 µm (Fig. 5D); YpsA-GFP: 3.34 ± 0.94 µm (Fig. 5G)

Identification of amino acid residues important for YpsA function 178
Aided by the crystal structure and computational analysis of the YpsA family of SLOG 179 domains we identified the conserved residues that are predicted to be important for 180 maintaining the function of YpsA ( Fig. 1B; see arrows). We performed site-directed 181 mutagenesis of two of these key residues and generated GFP-tagged ypsA variants 182 G53A and E55Q. We also generated other mutants to more generally explore YpsA 183 function namely, G42A, E44Q, W45A, W57A, or W87A. We ensured that all mutants were stably produced through immunoblotting (Fig. 6B). Microscopic examination 185 revealed that all YpsA variants except W57A were unable to trigger filamentation upon 186 overexpression (Fig. 6A), suggesting that YpsA function is compromised in all these 187 cases. We also noticed that G53A, E55Q, W45A, and W87A mutants displayed impaired 188 ability to form foci. This is consistent with the observation that the first two of these 189 mutations disrupt the conserved, predicted nucleotide-binding site of the YpsA family 190 (21), and the latter two likely disrupt a key strand and helix of the Rossmannoid fold. 191 192

Putative interaction partners of YpsA 193
To understand the role of YpsA via identifying its potential interaction partners, we 194 conducted FLAG-immunoprecipitation using  constructs as baits. Untagged YpsA served as our negative control. After confirming the 196 enrichment of proteins in the eluate fractions through silver staining and anti-Flag 197 immunoblotting, the samples were submitted for protein identification via mass 198 spectrometry. To identify proteins that specifically interact with YpsA, we eliminated all 199 proteins that also appeared in our negative control, as they are likely non-specifically 200 bound proteins and retained only proteins that were present specifically in both YpsA-201 FLAG and YpsA-GFP-FLAG eluates. A selective list of protein interaction partners of 202 YpsA is shown in Table.1. The entire list is provided in Table. S2. In addition to our bait, 203 FLAG-tagged versions of YpsA and presumably native copies of YpsA due to self-204 assembly, we noticed many proteins whose genes are under nutrient availability-sensing 205 CcpA (31) or CodY (32) or AbrB (33) regulon(s) in our IP results. Interestingly, several of 206 the proteins that associate with YpsA bind NAD or its derivatives and/or play a role in 207 redox-sensing. However, given that these are abundant metabolic enzymes we cannot 208 be sure of the significance of these interactions at this time. 209

Overproduction of YpsA inhibits cell division in S. aureus 211
To investigate if the role of YpsA is conserved in other Firmicutes, we chose to study the 212 function of YpsA in S. aureus. Cells lacking intact ypsA in S. aureus (34), are viable and 213 their cell morphology appear similar to WT control suggesting ypsA is not an essential 214 gene (Fig. 7AB). Next, we placed S. aureus ypsA (ypsA SA ) under the control of xylose-215 inducible promoter in a S. aureus plasmid vector. As shown in Fig. 6, the cell diameter of 216 Bacterial cell division is a highly regulated process and many division factors have 227 already been characterized especially in model organisms E. coli and B. subtilis. Yet, 228 cell division is only mildly affected even in the absence of a combination of known 229 division regulators in these organisms (12), thus predicting the presence of other 230 proteins that could affect the cell division process. Here, we discuss the role of YpsA, a 231 protein of hitherto unknown function conserved in diverse Firmicutes. We show that 232 YpsA offers protection against oxidative stress. However, the precise mechanism of how 233 this is achieved remains to be elucidated. Next, we show that YpsA overproduction leads 234 to impaired FtsZ ring assembly and ultimately cell division inhibition. 235 It has been reported that cotD-ypsA transcriptional unit is repressed by the regulator 237 essential for entry into sporulation, Spo0A (35), which binds to a region upstream of cotD 238 (36). It has been shown that cotD is also repressed by a late stage sporulation-specific 239 1B; (21)]. Since foci-formation was disrupted in both G53A and E55Q mutants, it is 257 plausible substrate-binding allows for multimeric complex formation. It is noteworthy that 258 mutants such as G42A and E44Q which are able to form foci, therefore likely bind 259 substrate, lack the ability to elicit filamentation. Also, YpsA-GFP overproducing cells 260 grown in CH medium were able to form foci but unable to induce filamentation (Fig. 5F). 261 These observations support a model in which substrate binding by YpsA is a prerequisite for cell division inhibition but substrate binding alone is not sufficient to 263 induce filamentation, assuming foci formation is indicative of substrate binding. It is Although the GxD/E motif is conserved in YoqJ, several residues we identified to be 284 essential in YpsA are not conserved in YoqJ ( Fig. 1B and Fig. 6). The Firmicutes-specific 285

Strain construction and general methods 305
All B. subtilis strains used in this study are isogenic derivatives of PY79 (49). See table 306 S1A for strain information. Overproduction of YpsA was achieved by PCR amplifying 307 ypsA using primer pairs oP106/oP108 (see table S1B for oligonucleotide information) 308 and ligating the fragment generated cut with SalI and NheI with IPTG-inducible amyE 309 locus integration vector pDR111 (D. Rudner) also cut with SalI and NheI and the 310 resulting plasmid was named pGG27. To construct a GFP fusion, ypsA fragment that 311 was amplified with primer pairs oP106/oP107 and digested with SalI and NheI was 312 ligated with gfp fragment generated with oP46/oP24 and cut with NheI/SphI and cloned 313 into pDR111 digested with SalI/SphI resulting in plasmid pGG28. The G42A, E44Q, W45A, G53A, E55Q, W57A, and W87A mutations were introduced using the 315 QuikChange site-directed mutagenesis kit (Agilent) using pGG28 as template. ypsA-316 3xflag was constructed via two step PCR using pGG27 as a template. Round one PCR 317 was completed using primers oP106 and oP291. The PCR product from round one was 318 then used as a template for round two PCR, which was completed using primer pairs 319 oP106 and oP292. The final PCR product was then cloned into pDR111 using SalI and 320 NheI restriction sites, making plasmid pRB33. Similarly, ypsA-gfp-3xflag was constructed 321 via two step PCR using pGG28 as a template. Round one PCR was completed using 322 primers oP106 and oP349. The PCR product from round one was then used as a 323 template for round two PCR, which was completed using primers oP106 and oP350. The 324 final PCR product was then cloned into pDR111 using SalI and SphI restriction sites, 325 making plasmid pRB34. The engineered plasmids were then used to introduce genes of 326 interest via double crossover homologous recombination into the amyE locus of the B. 327 subtilis chromosome. Expression of ypsA-his in BL21-DE3 Escherichia coli cells was 328 achieved by PCR amplifying ypsA-his with primer pairs oRB9 and oRB33, and cloning 329 into XbaI and BamHI resticition sites of pET28a, producing plasmid pRB21. YpsA-his 330 was purified using standard protocol involving nickel column-based affinity 331 chromatography. To produce S. aureus YpsA in S. aureus strain SH1000, ypsA SA 332 fragment (PCR amplified with oRB27/oP314 primer pairs) was cloned into xylose-333 inducible pEPSA5 plasmid using EcoRI and BamHI restriction sites (50), generating 334 plasmid pRB36. Plasmids were first introduced into S. aureus RN4220 via 335 electroporation, and then transduced into SH1000 (18). 336 337

Media and culture conditions 338
Overnight B. subtilis cultures grown at 22 °C in Luria-Bertani (LB) growth medium were 339 diluted 1:10 into fresh LB medium and grown to mid-logarithmic growth phase (OD600 = 0.5), unless otherwise stated. Expression of genes under IPTG-controlled promoter was 341 induced by addition of 1 mM IPTG (final concentration) to the culture medium unless 342 noted otherwise. Overnight S. aureus cultures were grown at 22°C in tryptic soy broth 343 (TSB) supplemented with 15 µg/ml chloramphenicol and/or 5 µg/ml erythromycin where 344 required for plasmid maintenance. Cultures were then diluted 1:10 into fresh medium 345 containing appropriate antibiotics and grown to mid-logarithmic growth phase (OD600 = 346 0.5), unless otherwise stated. Expression of genes under xylose-controlled promoter 347 was induced by the addition of 1% xylose when required. 348 349

Sporulation assay 350
Sporulation assay was conducted using resuspension protocol as described previously 351 (30). Briefly, overnight cultures of B. subtilis cells were grown in LB medium at 22°C, 352 were diluted 1:10 in fresh casein hydrolysate medium (CH, KD Medical) and grown to 353 mid-log phase twice before culture was resuspended in Sterlini-Mandelstam sporulation 354 medium (SM, KD Medical) to induce sporulation (51). Growth in CH medium and entry 355 into sporulation in SM medium were monitored via fluorescence microscopy. Total viable 356 cell counts (CFU/ml prior to heat treatment) and spore counts (CFU/ml after incubation 357 at 80°C for 10 min) were obtained for calculating sporulation frequency (spore 358 count/viable count). 359 360

Disc diffusion assay 361
All disc diffusion assays were completed on LB agar plates. Strains PY79 and RB42 362 were grown until OD600=0.5, and then 100µl of each culture was added to the respective 363 plates. Briefly, 15µl of 1M hydrogen peroxide was added to 7mm Whatman paper discs, 364 which were then placed equidistant from each other on top of the inoculated media. A 365 disc containing no hydrogen peroxide was used as a negative control. Plates were then 366 incubated overnight at 37°C. The diameter of the disc (7mm) was subtracted for the 367 zone of inhibition measurements. 368 369

Immunoprecipitation and mass spectrometry 370
The YpsA-FLAG immunoprecipitation was performed using 371 FLAGIPT1 immunoprecipitation kit (Sigma-Aldrich) as described previously (52) Table S1. Strains and oligonucleotides used in this study 528     A S R S E F A D L L K Q @ 5 V S E I D G N P T @ 5 K K A Y L R A G Q Q T V E C C @ @ D I L I A L W D G E @ 3 G T G G T A E I I A F A L Q @ R K K P V F I V S 185 consensus/80% .