Comparative genome analysis of Bacillus okhensis Kh10-101T reveals insights into adaptive mechanisms for halo-alkali tolerance

Background: Bacillus okhensis , isolated from saltpan near port of Okha, India, was initially reported to be a halo-alkali tolerant bacterium.We previously sequenced it’s 4.86 Mb genome, here we analyze its genome and physiological responses to high salt and high pH stress. Results: B. okhensis is a halo-alkaliphile with optimal growth at pH 10 and 5% NaCl. 16S rDNA phylogenetic analysis resulted in habitat based segregation of 106 Bacillus species into 3 major clades with all alkaliphiles in one clade clearly suggesting a common ancestor for alklaliphilic Bacilli. We observed that B. okhensis has been adapted to survive at halo-alkaline conditions, by acidication of surrounding medium using fermentation of glucose to organic acids. Comparative genome analysis revealed that the surface proteins which are exposed to external high pH environment of B. okhensis were evolved with relatively higher content of acidic amino acids than their orthologues of B. subtilis. It posess relatively more genes involved in the metabolism of osmolytes and sodium dependent transporters in comparison to B. subtilis. Growth of B. okhensis is Na + dependent, with a minimum requirement of 4% NaCl at neutral pH but 0.5% NaCl is enough at pH 10. It tolerated sudden increase of salt concentration of its medium, and exhibited an elongated cellular phenotype. But, could not tolerate a sudden shift of pH from 7 to 11, and cell envelope got damaged, conrming the pH regulation through cell wall reinforcement is key to survival at high-pH condition. We observed that hydroxyl ions damage the cell, but not Na+ ions, becuase at high pH Na+ ions were not accumulated inside. Conclusion: B. okhensis uses acidication of the external medium and pH dependent cell wall reinforcement to survive sodic environments. Interestingly, its growth is highly Na + dependent and the genome encodes for a high proportion of acidic amino acids in majority of surface proteins in comparison to their orthologues of B. subtilis, a direct evidence of adaptive evolution. comparison their direct wide comparison

Bacillus okhensis strain Kh10-101 T (here after B. okhensis) is a Gram-positive, strictly aerobic, rod like bacterium, rst isolated from a salt pan in India, near the port of Okha by Nowlan et al., (2006) [23]. In this work, we analyzed the physiological response of B. okhensis to high pH and high NaCl conditions together with a comparative genome analysis with other halo-alkaliphilic species B. pseudo rmus OF4, B. halodurans C-125, B. alcalophilus ATCC27647, B. clausii KSM-K16 and a neutrophilic, B. subtilis subsp. subtilis strain168 from a halo-alkali tolerance standpoint. The comparative genome analysis was performed to differentiate the high salt responsive mechanisms from the high pH mechanisms and the hydroxyl ion (OH − ) effects from sodium (Na + ) toxicity. Our ndings reveal the insights into halo-alkali tolerance mechanisms and emphasize for the rst-time hydroxyl ion stress in high pH condition other than sodium toxicity.

Results
B. okhensis is a halo-alkaliphile B. okhensis was originally identi ed as a halo-alkali tolerant bacterium [23], in our laboratory conditions optimal growth was observed when B. okhensis was allowed to grow in CMB medium with pH adjusted to pH 10 and 5% NaCl (Fig. 1A). Therefore, pH 10 and 5% NaCl concentration were used in all experimental conditions unless otherwise speci ed. Growth was observed with varying growth rates in NaCl concentrations ranging from 0.5 to 12% at optimal pH (pH 10). Similarly, growth was observed in media with a pH ranging from 7 to 11, when grown in the optimal NaCl concentration (5% NaCl). When the culture was grown in the medium containing a high salt (12 % NaCl) concentration and optimal pH, it exhibited an extended lag-phase (Fig. 1A). In contrast, when grown at low salt concentration (0.5% NaCl) at optimal pH the culture reached to stationary phase in ~6 h and a rapid decline in the cell density was observed afterwards, clearly demonstrating requirement of NaCl (5%) in the medium for stable cell density (Fig. 1A). Whereas, when cells were grown at a neutral pH (pH 7) and optimal salt concentration, nal cell density was relatively low compared with that of the optimal growth conditions ( Fig. 1A). A lag phase of 4-5 h was observed at high pH (pH 11) and optimal salt concentration with a slower growth rate ultimately reaching a cell density similar to that of optimal conditions during stationary phase (Fig. 1A).
The data suggests that B. okhensis is a moderate halophile with minimum salt being essential for survival and optimal growth at 5% of NaCl. Interestingly it is also an alkaliphile with high nal cell density at alkaline pH (pH 10) when compared with neutral pH growth conditions (Fig. 1A).
Cell morphology was examined using SEM imaging in the same growth conditions as mentioned above with cells were collected during the logarithmic growth phase. Cells grown at optimal conditions exhibited an average cell length of 3.5±0.6 µm with an average cell thickness of 0.87µm (Fig. 1B&C). When cells were grown in neutral or at high pH conditions, there was no signi cant difference in the cell morphology with an average cell length of 3.49±0.6 µm and 3.7±0.9 µm respectively (Fig. 1B, D&E). However, when grown in low salt concentration at an optimal pH, cells were smaller than that of the cells grown in optimal growth conditions, with an average cell length of 2.7±0.6 µm (Fig. 1B&F). In contrast, cells grown in high salt at the optimal pH exhibited an elongated phenotype with an average cell length of 5.4±1.5 µm (Fig. 1B&G).

Phylogenetic analysis of B. okhensis
In order to identify the phylogenetic relationship of B. okhensis with other extremophilic Bacilli, we have generated a phylogenetic tree using 16S rRNA maximum likelihood phylogeny. 16S rDNA, bootstrap consensus phylogenetic tree of 106 Bacillus spp., has exhibited 3 major clades (Fig. 2) (Fig. 2).
The largest clade with 72 species contained all neutrophilic Bacilli. It is important to note that segregation of alkaliphiles and neutrophilic Bacilli con rms evolution of alkaliphiles from a possible common ancestor indicated with an arrow (Fig. 2). Bacillus okhensis contains genes for acidi cation of external environment and reduces the external medium pH: Changes in the pH of external medium and growth of B. okhensis were monitored together after inoculation of the culture into the medium having different combinations of pH and NaCl concentration.
Interestingly, in CMB medium with glucose as carbon source, the pH of the external medium decreased with an increase in cell density, irrespective of the starting pH of the growth medium. When cells were inoculated at optimal conditions (pH 10 and 5% NaCl) in CMB medium, within 60 min, pH of the medium was reduced to ~9.5 with a simultaneous growth of the cultures and reached to pH 6 at stationary phase (Fig. 3A). However, when the culture was inoculated in a CMB medium at pH 11, no increase in the cell density was observed up to ~5 h after inoculation. During this long lag phase, pH of the medium got reduced gradually to ~pH 9.5 (Fig. 3B). Once the medium pH approached 9.5, the cell density started increasing. This indicates that when inoculated at high pH conditions, cells could be releasing organic acids to reduce the external medium pH before they actively start dividing. Interestingly, when inoculated at pH 7, the culture simultaneously started dividing with a concomitant decrease in the pH of the medium reaching to pH 6 in ~7 h post inoculation. However, the culture quickly reached stationary phase without producing much biomass; as the pH decreased further below pH 6.5 (Fig. 3C). The glucose consumption of cells grown in a high pH medium (pH 11) was monitored along with changes in the external pH of the medium. The amount of glucose decreased over time and reached a steady state after ~7 h inoculation (Fig. 3D). The decrease in the glucose levels in the external medium appeared to mimic the decrease in the pH of the medium, which reached pH 6.8 at stationary phase after 24 h (Fig. 3D). It is likely that the external medium pH was reduced by fermentation of glucose to acidic components. This was further con rmed by replacing glucose with non-fermentable malate as a carbon source. When cells were inoculated in a malate containing medium with an initial pH 9, no signi cant decrease in the pH of external medium was observed (Fig. 3E). Whereas with glucose, the pH of the medium was reduced to 6.2 from pH 9 after 24 h of incubation, con rming glucose fermentation is the source of the acidi cation of external medium (Fig. 3E).
The B. okhensis genome was compared with B. subtilis for genes involved in fermentation of carbohydrates to different organic acids, as it is clear from our data that the B. okhensis uses glucose to decrease the external pH of the medium. RAST subsystem features for fermentation and organic acid metabolism were presented in Table 2. The genome of B. okhensis encodes genes for metabolic pathways involved in the synthesis of organic acids, acetolactate, lactate, butyrate, acetoin, butane-diol and butanol. However, fermentation to acetolactate, butyrate and lactate were common in all the genomes compared in this study including B. subtilis (Table 2). Citrate metabolism and alpha acetolactate operon were observed to be present only in B. subtilis but not in alkaliphiles. The fermentation to mixed acids pathway is present only in B. okhensis and B. alcalophilus but not in the genome of B. subtilis and other alkaliphiles. The presence of fermentation to mixed acid pathway in B. okhensis makes it interesting as the pH of the growth medium is known to in uence the type of fermentation product produced [25]. Thus, an active mixed acid pathway could be an important high pH responsive mechanism in B. okhensis and B. alcalophilus (Table S1). In B. okhensis with a predicted formate e ux transporter (WP_034625882.1), formic acid seems to be a potential exudate involved in pH reduction along with lactate, acetolactate and butyrate.
Adaptation of proteins for survival at extreme alkaline conditions in B. okhensis: The pI and acid to base ratio of all the proteins from B.okhensis and B.subtilis were compared to analyse the changes that could have taken place during the course of adaptive evolution to cope up with the alkaline pH (Fig. 4). It is interesting to note that irrespective of their survival at alkaline or near neutral pH conditions, none of the proteins have pI near the biological pH, 7.5 in these two Bacillus species (Fig.  4A&B grey highlighted). Though B. okhensis thrives at alkaline habitat, it seemed to maintain the cytoplasmic pH close to near neutral pH, as none of the protein's pI value is in the range from pH 7 to 7.5. Majority of its cytosolic proteins have acid to base ratio and pI values comparable with B. Subtilis (Fig. 4A). Over all there is a slight shift in the pI curve of proteome of B. okhensis towards acidic region when compared to B. subtilis (Fig. 4B). Similar pattern of pI shift of proteins towards acidic region and lack of proteins with pI near probable cytoplasmic pH was observed in all alkaliphiles used in this study (Fig. S1). There are many proteins with pI in the acidic range in alkaliphile and alkaline range in B. subtilis (indicated with open and closed arrows respectively in Fig. 4B). Interestingly, about 122 proteins of B.okhensis showed signi cantly lower pI and acid to base ratios when compared to the corresponding orthologs from B. subtilis (Table S2, Fig. 4C). It is interesting that almost all the proteins with relatively low pI values in B. okhensis, but not in B. subtilis were surface proteins such as, cell wall, outer membrane, agellar, spore formation related proteins and some hypothetical proteins. This demonstrates that, as these surface proteins are often being exposed to high pH and affected by the external environment than the cytosolic proteins, B. okhensis might have evolved with relatively higher acid to base ratio in the proteins during the course of adaptive evolution.
Role of cell envelope in halo-alkali tolerance Genome of B. okhensis was searched for the presence of homologous genes of cell wall synthesis, which were reported to play an essential role in the pH homeostasis in B. halodurans C-125 [14,26]. We found that the B. okhensis genome has 7 Teichuronic acid biosynthesis genes, whereas extreme alkaliphile B. pseudo rmus, which can tolerate sudden shifts from low to high pH has only 3 gene (Table 3). It has been reported that constitutively expressed genes for formation of S-layers contribute to the adaptability to sudden shifts to extreme alkalinity but are unfavorable to the growth of B. pseudo rmus OF4 at nearneutral pH [1]. Interestingly, RAST SEED has more orthologs and subsystem features annotated for genes responsible for capsular and extra cellular polysaccharide synthesis in neutrophilic B. subtilis than extreme alkaliphiles (Table 3). Surprisingly, genes coding for Murein biosynthesis integral membrane protein MurJ (WP_034630126.1) and MFS transport protein, YceL (WP_034626199.1) were identi ed in B. okhensis but not in any of the compared strains. In alkaliphiles the head group of the phospholipids in the membrane often consists of an array of branched chain negatively charged lipids, such as cardiolipin, phosphatidylglycerol, and phosphatidylethanolamine. Of these, cardiolipin is the most important component contributing to pH homeostasis. A high concentration of cardiolipin in the plasma membrane is a characteristic feature of alkaliphilic Bacilli [27]. In the genome of B. okhensis, three genes coding for cardiolipin synthase (Cls), WP_034629090.1, WP_034631834.1 and WP_034631890.1 were identi ed. It is interesting to note that the B. okhensis genome harbors a pair of genes that encodes DesA (fatty acid desaturase WP_034627982.1) and acyl-CoA desaturase (WP_034628144.1). Even the genome of the extreme alkaliphile, B. pseudo rmus encodes a gene for fatty acid desaturase (WP_012957426.1).
However, it has been reported that the facultative alkaliphiles lack desaturase activity and incorporation of unsaturated fatty acids in the membrane restricts these organisms' ability to grow at near-neutral pH [28].
Adaptations for high salinity Uptake of potassium (K + ) plays an important role, along with osmo-protectants such as glycine betaine, in salt stress tolerance [29,30]. The B. okhensis genome contains genes coding for ve different potassium channels, which include 4 potassium channels (WP_034633860.1, WP_034632084. . It has highest number of genes related to the glycine betaine transport when compared to the other species suggesting that it also relies on chemical osmo-protectants to thrive under extreme halo-alkaline conditions ( Table 4).
The B. okhensis genome harbors several genes related to the production of sugar alcohols, such as mannitol- . Thus, the B. okhensis genome has evolved with genes responsible for adaptation to high salt in its natural habitat, saltpan. The number of genes annotated and predicted to be involved in synthesis of different osmolytes and ectoine in different Bacillus strains is presented in Table 4.

Sodium dependent growth and survival of B. okhensis and pH homeostasis
Growth of B. okhensis in CMB medium with different sub-optimal NaCl concentrations was monitored. B. okhensis required a minimum of 4% NaCl to survive at neutral pH 7. When inoculated in a medium containing low salt (i.e., 0.5 % NaCl) it is necessary to set a high pH medium for its growth, suggesting sodium ion dependency for its survival (Fig.5). Without NaCl in the medium, growth was not observed at any pH (7 to 10) (Fig. 5A). However, when inoculated in a medium with a pH of 9 or 10 in the absence of NaCl, the maximum cell density obtained was very low and the same was decreased immediately showing absolute NaCl requirement for survival of B. okhensis ( Fig 5A). With 0.5% NaCl in the medium, cells exhibited signs of growth at pH 10, but not in pH 7, 7.5, 8 or 9 (Fig. 5B). A slight increase in the cell density was observed at pH 9 but this was signi cantly lower than that of pH 10 ( Fig 5B). Interestingly, even at pH 10 and 0.5% NaCl the cell density decreased after 7h of growth suggesting cell lysis due to insu cient Na + /NaCl availability (Fig. 5B). In the presence of 2% NaCl in the medium, cells exhibited growth at pH 8, 9 and 10 but not at pH 7.5 or pH 7 (Fig. 5C). At a concentration of 2% NaCl, cell density was stable at pH 10, but not in pH 9 or pH 8 ( Fig 5C). In the medium with 4% NaCl cells started growing in all the pH conditions tested, with a stable cell density being seen during stationary phase (Fig. 5D). These observations con rmed that a certain amount of Na + ions are essential for the growth and survival of B.
okhensis. This sodium ion dependency of growth is expected to be due to the role of Na + in nutrient transport as well as high pH tolerance.
Since we observed that the sodium ions play a key role in growth of B. okhensis, we analysed the genome of B. okhensis for candidate genes involved in sodium dependent transport. Extreme alkaliphilic Bacillus strains are generally believed to rely heavily on dedicated, highly effective Na + /H + antiporters by which, H + ions from the external environment are transported into the cell in exchange for sodium ions (Na + ), thus acidifying the cytoplasm, for growth and survival in alkaline habitats. The main mechanism of pH homeostasis in B. okhensis is probably catalysed by NhaC and NhaD antiporters (WP_034627995.1and WP_034625225.1). Interestingly, NhaD in B. okhensis has an ortholog in extreme alkaliphile B. pseudo rmus (84% identity) but not in other species compared. B. okhensis genome has 12 genes coding for Na + /H + antiporter subunits which is similar in other Bacillus species (Table 4). In sodic environments cells require more energy to survive as they must expend more energy taking up solutes and extruding sodium ions [2,20]. The requirement for robust Na + extrusion is most likely important because most of the ion-coupled solute symporters and agellar motors of B. okhensis are Na + coupled. Indeed B. okhensis has 10 candidate genes predicted to be involved in sodium dependent transport of nutrients and 8 candidate genes involved in sodium-based symport of nutrients, whereas in B. subtilis only 4 were present ( Table 4). Molecules that are transported in a sodium ion dependent manner include glutamine, glutamate, proline, phosphate, sulphate, bicarbonate and alanine. This family of transporter proteins function to exchange internal sodium ions with external solute molecules or these solutes are cotransported into cell using a sodium gradient. Moreover, the genome of B. okhensis has a sodium ion in ux: the voltage-gated sodium channel (WP_052144823.1) and the sodium symporters and transporters (22 proteins). B. pseudo rmus OF4 mutants with no functional NaVBP fail to maintain pH homeostasis during sudden alkaline shifts in the external pH [32]. The NaVBP sodium channel of B. okhensis is closely related to that of B. wakoensis (81% sequence similarity) and exhibited no signi cant similarity with that of B. pseudo rmus, B. halodurans and B. alcalophilus. Na + re-entry in alkaliphiles can happen through the Na + coupled agellar Mot complex, which functions as an ion channel while converting chemiosmotic energy into mechanical energy. The motPS genes coding for Na + coupled Mot complex are mostly found in alkaliphiles, such as B. pseudo rmus OF4, B. halodurans C-125, and O. iheyensis HTE831 [32]. Whereas, proton coupled Mot complexes coded by motAB genes are widely distributed among Bacillus spp., including the alkaliphiles. The genome sequence of B. okhensis revealed only a single set of genes encoding a MotAB-like pair of proteins as the stator ( agellar motor protein MotP, WP_034629847.1 and agellar motor protein MotA WP_034630220.1). These stator proteins are closely related to that of B. wakoensis and B. akibai with similarity of 86% and 85% respectively but not shown any signi cant similarity with that of B. pseudo rmus or with B. halodurans and has exhibited 36% identity with the MotA in B. clausii, which is known to change coupling ion from a proton to a sodium ion in response to external pH [32].
Response of B. okhensis to a sudden shift to high salt Bacteria can cope up with a sudden increase in salinity by activating acclimation mechanisms. To investigate the effect of exposure to high salt sudden shift on the growth and cell morphology of B. okhensis, cells in exponentially growing phase at low salt (OD 520 , 0.2) were shifted to high salt condition.
Initially no increase in the cell density was observed for a period of 60 min after a sudden shift from 0.5 to 12% NaCl and then growth had resumed (Fig. 6A). As already mentioned, B. okhensis grows to low cell density without any stable stationary phase (Fig. 1A). But, shift of low salt (0.5%) acclimated cells to high (12%) salt led to higher cell density with stable stationary phase. The cell morphology of these high salt treated cells were examined using SEM. Cells were taken at 30, 60 and 180 min after the shift to 12% salt and compared to an untreated reference (Fig. 6B). Cells exhibited an elongated phenotype during high salt treatment (Fig. 6C, D and E). Within 30 min no signi cant difference in the cell size and shape was observed. After 60 min of high salt treatment, an increased number of cells exhibited elongated phenotype (Fig. 6D). After 1 hour of treatment, the cells began to divide resulting in a mix of elongated and short cells, con rming that some of the elongated cells started dividing after acclimation to high salt (Fig. 6E). In previously studied non-halotolerant Bacillus models, salt stress resulted in formation of lamentous cells due to inhibition of cell septum formation. This lamentous phenotype was attributed to inhibition of autolysins, involved in hydrolyzing peptidoglycans, which allows the daughter cells to separate during division [33]. Similarly, inhibition of cell septum formation might be involved in elongated phenotype in B. okhensis in response to high salt stress that was also observed in B. cereus [34]. In addition, putative cell division inhibitor (yfcH homolog) gene was identi ed (WP_034626153.1) in the genome of B. okhensis. Our observations suggest that sudden shift from low salt (0.5%) to high salt (12%) has not shown any lethality or signi cant damage of metabolism or cell growth and survival. It clearly indicates that the genome of B. okhensis has evolved with mechanisms to quickly adapt to sudden changes in external NaCl concentration.
High pH shock and Hydroxyl ion stress Bacterial cells must maintain near-neutral cytoplasmic pH and devoid of sodium toxicity to survive in alkaline-sodic environment [2,20,22]. A sudden shift in pH from low to high will create alkaline shock or hydroxyl ion stress on the cells. It was reported that B. subtilis showed no growth following a rapid pH increase [35]. B. okhensis cells those were pre-adapted to pH 7 were given a high pH shock by changing the pH of the medium to pH 11. This sudden shift in pH is followed by no increase in cell density for several hours (Fig. 7A), growth resumed after ~36 h of pH shock (results not shown). Even though, B. okhensis is a halo-alkaliphile and has evolved with adaptive mechanisms to tolerate high pH during the course of evolution, cells which were fully acclimated to pH 7 cannot tolerate the change to a high pH. It is likely that the organism might have turned off the adaptive mechanisms needed for growth under high pH conditions. After the shift of cultures from pH 7 to 11, the bacterial cells became wrinkled and severely damaged (Fig. 7B, C&D, Fig. S2). Such changes were not observed when pH 10 acclimated cells were shifted to high pH 11 (data not shown). This data clearly indicates that the all the mechanisms required for survival at pH 10 and above are operative during their growth, therefore a shift from pH 10 to 11 did not cause any cellular damage. When the culture from optimal growth conditions (pH 10 and 5% salt) was inoculated to pH 7 after several generations, it might have acclimatized to survive at pH 7, perhaps by turning off all high pH-tolerant mechanisms. Inoculation into pH 11 medium, from pre-cultures maintained in pH 9 or 10 mediums does not shown damage to the cell surface (Fig. 1E). This suggests that an alkaliphile could adapt to near neutral pH condition by switching off the mechanisms that are necessary for high pH tolerance. In B. subtilis transport of inorganic ions, polyamines and several other genes were upregulated upon high pH stress, which includes K + /H + antiporters, sig-ω regulon [35]. Similar mechanisms might exist in B. okhensis. Indeed, upon a high pH shift both the B. okhensis membrane and cell wall were damaged.
Change in cellular Na + /K + , with salt stress and high pH stress As observed in previous sections, a sudden change in the pH of the medium with NaOH caused severe cellular damage, but the same was not observed with high NaCl shock. This clearly indicates that at a high pH, cells have to deal with hydroxyl ions along with sodium toxicity. To differentiate the damage made to the cells by hydroxyl ions from that of sodium cytotoxicity, entry of sodium ions into the cells due to high pH challenge was evaluated. Cellular elemental composition was estimated using SEM assisted EDX (Fig. 8). SEM assisted EDX analysis has shown that the cellular content of B. okhensis grown in optimal conditions contained 32.9% Na + and 5.48% K + (Na + /K + ratio 6) respectively. When grown at a neutral pH the Na + /K + ratio was 0.91(13.8/15.1% relative units respectively). At 30 min after shifting to high pH the ratio increased to 16.09 (35.4/2.2). This clearly indicates a sudden shift in pH leads to increased cellular Na + ion concentration with a concomitant decrease in K + concentration. This rise could be because of damage to the membrane or an increase in Na + ions inside the cell due to change in equilibrium of the cellular sodium cycle. However, when cells were allowed to grow at 0.5% NaCl the Na+ / K+ ratio was 1.15 (14.4/12.5% relative units respectively). But at 30 min after a shift of cultures from low salt to high salt concentration (0.5 to 12%), the ratio was increased to 2.89 (34.2/11.8). From these observations, we conclude that cells accumulated similar amount of sodium ions, when shifted from 0.5 to 12% NaCl as well as upon a sudden shift from pH 7 to 11. This could be due to the damage caused by the excessive hydroxyl ions to the cell wall as observed under SEM (Fig. 6). It is not the entry of sodium ions or sodium toxicity that caused the cellular damage during high pH stress, because EDX analysis has shown that shift to high pH led to no signi cant increase in the internal sodium ions in comparison with cells grown in optimal conditions. Therefore, the physical damage to the cell wall could be due to hydroxyl ion stress.

Genes coding for enzymes with potential applications in industry
Various alkaline enzymes, such as proteases, amylases, and cellulases, have been successfully commercialized on an industrial scale with a signi cant share in global enzyme industry [7,36]. We

Discussion
The natural cause of sodic environments is the presence of soil minerals producing sodium carbonate (Na 2 CO 3 ) and sodium bicarbonate (NaHCO 3 ). Therefore, saline stress and sodium toxicity overlaps with high pH stress in alkaliphilic Bacillus. B. okhensis reported as halo-alkalitolerant bacterium; is indeed a moderate halophile and an alkaliphile with growth optima falling at 5% NaCl and pH 10 in CMB medium. B. okhensis upon incubation in 12% NaCl exhibited elongated phenotype explaining probably high salt stress resulted in inhibition of daughter cells partition during cell division ( Fig. 1G and Fig. 6D&E). Based on 16S rDNA sequence similarity B. okhensis is closely related to B. wakoensis and there is a possible common ancestor for all the alkaliphilic Bacillus spp. (Fig. 2). Interestingly B. okhensis has a relatively large genome with more CDS regions than its closest alkaliphilic models (Table 1). Genes for fermentation to organic acids was observed in all the species compared, however formate e ux is a possible high pH response observed in B. okhensis (Table 2&S1). Along with genes predicted for fermentation of glucose to organic acids acidi cation of external medium was demonstrated in B. okhensis which was further con rmed by using malate as carbon source ( Table 2, Fig. 3A, B, C, D&E). Based on earlier reports of Bacillus cell wall composition and its relation to tolerance of high pH, no correlation was observed with pH tolerance and the number of genes predicted to be involved in cell wall synthesis or membrane composition. Extreme alkaliphilic bacterium B. pseudo rmus, or B. okhensis has only Teichoic acid (TA) biosynthesis and Teichouronic acid biosynthesis genes but alanyl-lipoteichoic acid biosynthesis, sortase metabolism and polyglycerolphosphate and lipoteichoic acid metabolism genes are not detected (Table 3). Contrasting from previous reports from Dunkley et al., (1991) [28], a membrane desaturase gene was identi ed in B. okhensis and even in extreme alkaliphile B. pseudo rmus. Further biochemical and mutational studies are required to understand the role of membrane composition, cell wall and exo-polysaccharide role in high pH tolerance. B. okhensis harbors an evolved genome, well prepared to survive sudden shifts in NaCl concentration with a signi cantly higher number of genes involved in biosynthesis and transport of osmolytes in comparison with its nonhalophilic neutrophilic relative B. subtilis (Table 4). Our analysis reveals that irrespective of their pHenvironment, these bacterial species maintain their cytoplasmic pH close to 7 to 7.5 (Fig. 4A&B).
Strikingly, B. okhensis genome has evolved in such a way that the membrane bound, agella, spore formation related and other proteins being exposed constantly to external high alkaline environment contained higher acidic to basic amino acid ratio with lower pI values (Table S2, Fig 4C). Evolution of B. okhensis to its natural habitat, a saltpan was further con rmed by its tolerance to salt shift from 0.5% to 12% of NaCl ( Fig 6A). Sudden shift from low salt to high salt led to elongated cell phenotype and onehour lag phase but neither the growth nor the viability was in uenced. Interestingly, the B. okhensis cell envelope is unstable in low salt, cells lose viability with time. This was observed as a decrease in cell density over time (Fig 5-•-& Fig 1-○-) indicating requirement of NaCl for stable cell density and supports its halophilic nature. The presence of Na + dependent transportation and presence of more predicted candidate proteins involved in sodium cycle (Table 4) act as an evidence for the genetic makeup for survival at high Na + . Despite the presence of such sodium cycle and strong Na + /H + antiporters, cells of B. okhensis couldn't tolerate a sudden shift in external pH from pH 7 to pH 11 (Fig. 7). We also demonstrated that the pressure created by high pH physically damaged the cells and signi cant proportion of them were dead with porous membrane (Fig. S2, Fig 7D). It clearly indicates that cells which can tolerate higher Na + in the medium need not tolerate increased OHin real molar concentration i.e in salt shift (0.5% to 12%), increase in NaCl/ Na + is ~ 2M and in pH shift from pH 7 to 11 theoretical increase in Na + and OHis only 1mM. More interestingly EDX analysis indicated Na + were not accumulated in the cells due to high pH shift (Fig. 8). Similar results were observed with salt shift, clearly indicating the damage to the cell wall and loss of cell viability during high pH shift is due to OHion and charge imbalance (Fig. 8, Fig. 7D & Fig. S2).

Conclusions
B. okhensis is indeed a moderate halophile and an alkaliphile with optimal growth conditions of pH 10 and 5% NaCl. Our comparative genome analysis and physiological studies have shown B. okhensis uses acidi cation of the external medium and pH dependent cell wall reinforcement to survive alkaline environments. It can survive a sudden shift in NaCl concentration from 0.5% to 12% with elongated cell phenotype. In contrast, a shift from pH 7 to pH 11 has severe effect on cell growth and viability. Our observations have shown that during high pH there is hydroxyl ion induced damage on cell wall and membrane. The genome of B. okhensis has a relatively high number of genes involved in the metabolism of osmolytes and sodium dependent transporters in comparison to B. subtilis. Ectoine biosynthesis is observed only in alkaliphilic Bacilli but not in B. subtilis, explaining the adaptation of these alkaliphiles to high salt and sodic environments. Growth of B. okhensis is Na + dependent, with a minimum requirement of 4% NaCl at neutral pH but 0.5% NaCl at pH 10 due to Na + dependent transport of essential nutrients. B. okhensis Genome encodes for a high proportion of acidic amino acids and majority of proteins have low pI values in comparison to their orthologues of B. Subtilis. This is a direct evidence of adaptive evolution of organism to sodic/halo-alkaline environments. Genome sequence of B. okhensis has also opened a new direction of considering its industrial and biotechnological importance. Although the initial interest arose from the halo-alkali tolerance capacity of this organism, this study has revealed that the genetic content of B. okhensis and the other alkaliphilic Bacilli species have majorly evolved from the sodic environments and for sodium toxicity. Further to differentiate the sodic/Na + stress from that of high pH/OHstress and to understand the evolution of alkaliphiles, extensive genome wide comparison studies are needed to delineate salt stress and alkaline stress in halophiles, non-halotolerant alkaliphiles and halo-alkaliphiles.

Strain and culture conditions
Bacillus okhensis Kh10-101 T was obtained from Japan Collection of Microorganisms (JCM) Tsukuba, Japan. Growth was monitored in CMB medium [23] with slight modi cations, such as adjustment of medium pH with 10 M NaOH instead of Na 2 CO 3 after sterilization. For growth measurements cells were grown in 25 ml CMB medium in 100ml culture asks with side arms. An actively growing starter culture grown at pH 9 and 5% NaCl was used to inoculate experimental cultures. Cell density via turbidity was measured in a calorimeter with the lter set to 520 nm. pH of the medium was measured by centrifuging 50 ml of cells at respective time point at 8000 × g for 2 min at room temperature and the supernatant was used to measure the pH of the medium. Glucose estimation was done by colorimetric method [37]. For pH reduction using non-fermentable sugar, cells were grown in mineral medium with either glucose (1%) or with malate (1%) as carbon source with initial pH of 9 at the time of inoculation.
Low pH to high pH shift and low salt to high salt shift Cells were acclimatized to pH 7 and 5% NaCl on CMB agar plates for three subcultures. Cells collected from third subculture was inoculated in to 25 ml CMB liquid medium (pH 7 and 5% NaCl concentration).
Mid logarithmic phase culture was used to further inoculate into CMB medium with pH 7 and 5% NaCl. To give high pH shock (pH 11), 10M NaOH solution was added to early log phase culture. Cell viability after 30 min of high pH shock was veri ed using SYTOX green as per Krishna et al., [38]. Cells are not viable upon repeated subculture at 0.5% hence for salt shift experiment, log phase cells growing in pH 10 and 0.5% NaCl were used as starter culture and inoculated into pH 10 and 0.5% NaCl medium and at designated point of time (2 h), cells were added to a ask with autoclaved NaCl so that the nal concentration of NaCl is 12% in the external medium.

SEM microscopy
One milliliter of cells in the designated time point were pelleted (12000 x g, for 1 at 4 o C) and cell pellet was re-suspended in 1 ml of Karnovsky's xative (5% glutaraldehyde, 4% paraformaldehyde in 0.08 M sodium phosphate (pH 7.2) buffer and incubated on ice for 30 min. Cells were then harvested (12000 x g, for 1 at 4 o C), and cell pellet was re-suspended in 1ml fresh Karnovsky's xative and stored at 4 o C. For scanning electron microscope (SEM) imaging, the water in the cells was replaced with ethanol using gradient washes, from 10% to 100% with 10% increase in ethanol each round. Water replaced cells were gold sputtered and studied under scanning electron microscope (SEM). For cell length measurements, a minimum of 20 randomly selected cells were measured and average is represented as mean ± SD.
SEM-EDX measurements, 1 ml of culture was used to pellet the cells (12000 x g, for 1 minute at 4 o C) and the pellet was washed with 1ml of milliQ water by quickly re-suspending and centrifuging at 12000 x g, for 1 minute at 4 o C. The washed cell pellet was spread on stubs designated for EDX analysis, a unit area on cell was examined for Na, Mg, P, S, Cl, K, Ca, Mn, Fe, Co, Cu, Zn, elements; total signal was calculated to 100% and relative quantity of individual ions was measured.

16S rDNA phylogenetic analysis
Bacillus spp with reported whole genome sequences (Jan 2016) were selected for phylogenetic analysis. The evolutionary history was inferred by using the Maximum Likelihood method based on the Tamura-Nei model [39]. The bootstrap consensus tree inferred from 500 replicates [40] is taken to represent the evolutionary history of the taxa analyzed [41]. Branches corresponding to partitions reproduced in less than 50% bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (500 replicates) are shown next to the branches [40]. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. The analysis involved 17 nucleotide sequences. All positions containing gaps and missing data were eliminated. There was a total of 1011 positions in the nal dataset. Evolutionary analyses were conducted in MEGA7 [41].

Genome annotation and comparative analysis
Annotation was done using RAST tool and the number of representative genes coding for each subsystem were represented as per the RAST seed subsystems [42]. indicated in tables and corresponding protein id from NCBI was presented in text. All the predicted proteins from the RAST annotated, predicted proteomes were submitted to online tools http://web.expasy.org/cgi-bin/compute_pi/pi_tool for pI calculation.
Identi cation of bidirectionally best hits and comparison of their pI values with acidic to basic amino acid composition: To nd the orthologs between B. okhensis and B. subtilis the protein databases were created with the stand alone version of BlastP downloaded from the NCBI website ("ftp://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/LATEST/"). The fasta format le was generated for both B. okhensis and B. subtilis. These les having the protein sequences in fasta format were given as input and protein databases of both Bacillus species were created in Blast+. For identifying the true orthologs by bidirectionally best-hit method all the proteins of B. okhensis taken as query sequences were subjected to Blastp against the B. subtiius protein database as target. Similarly, Bacillus subtilis proteins were taken as query sequences and B. okhensis protein data base as target. The percentage identity, Evalue and the bit score were taken as parameters to choose the best hit. The normalized frequencies of all 20 amino acids of each and every protein were calculated using RStudio and the pI values were calculated. For every protein, acidic to basic amino acids' ratio (D+E/K+R+H) was calculated by adding the acidic residues and dividing it by the sum of the basic residues and a graph was plotted between pI values of every orthologous proteins and their corresponding Acid to base ratios.    okhensis grown at different combinations of pH and NaCl: pH 10 and 5% NaCl (C), pH 7 and 5% NaCl (D), pH 11 and 5% NaCl (E), pH 10 and 0.5% NaCl (F), pH 10 and 12% NaCl (G). Scale bar = 5 μm.  Reduction of external pH as an adaptation method; the pH of the external medium and cell density that were inoculated (pH 9 and 5% starter culture) into different combinations of initial pH and NaCl. pH 10 and 5% NaCl (A), pH 11 and 5% NaCl pH 7 and 5% NaCl (B), pH 7 and 5% NaCl (C), growth and pH of external medium was measured in each condition. Similar results were obtained in three independent experiments and the data represented as means ± SD. Consumption of glucose from the external medium and corresponding pH (D); Glucose concentration in the external medium (-•-), pH of the external medium (-○-). Cells adapted to pH 11 for 3 subcultures were used as the initial inoculum, decrease in glucose concentration and pH of the medium were monitored. Similar results were obtained in three independent experiments, and the data represented as means ± SD. Changes in the external medium pH in minimal medium (E). Fermentable carbon source glucose (-•-) and non-fermentable carbon source malate (-○-), similar results were obtained in two individual experiments and data represented is the mean.  Table S2.