Extracytoplasmic function (ECF) σ factors are a group of bacterial σ factors that direct transcription of genes involved in tasks such as maintenance of cell surface integrity in response to environmental stresses. Bacillus subtilis has seven ECF σ factors: σM, σV, σW, σX, σY, σZ and σYlaC. Except for σZ, these are regulated directly by their respective cognate transmembrane anti-σ factors, which sequester the σ factors and keep them inactive. Under stress conditions, the σ factors are released from the anti-σ factors and bind RNA polymerase core enzyme to form holoenzymes, which transcribe the respective regulon genes. The genes for the σ factors and the anti-σ factors form operons whose transcription is directed by the cognate σ factors; the activities of the promoters for ECF σ factor genes are thus indicative of the activities of the respective ECF σ factors    .
Among these ECF σ factors, σV primarily responds to lysozyme  . It is activated by stepwise proteolytic destruction of the anti-σ factor RsiV via regulated intramembrane proteolysis (RIP)  . Lysozyme directly binds to RsiV and induces the cleavage of its C-terminal extracytoplasmic portion by signal peptidase (site-1 cleavage). Subsequently the N-terminal portion is processed by the intramembrane protease RasP (site-2 cleavage), and σV is released   .
The B. subtilis membrane contains the glucolipids (monoglucosyldiacylglycerol, diglucosyldiacylglycerol and triglucosyldiacylglycerol) in addition to phospholi- pids (phosphatidylethanolamine, phosphatidylglycerol, cardiolipin and lysylphosphatidylglycerol)   . Glucolipids are synthesized by the gene product of ugtP, which processively transfers glucose from UDP-glucose to diacylglycerol  . UgtP protein is implicated in nutrient-dependent cell size control  . ugtP null mutants which completely lack glucolipids have been found to show abnormal cell morphology   and constitutive activation of three ECF σ factors σM, σV and σX   .
When the ugtP gene with a hexahistidine tag-coding sequence was placed under the control of the isopropyl-β-d-thiogalactoside (IPTG)-inducible Pspac pro- moter and integrated into the chromosomal aprE locus in the ∆ugtP mutant, the activities of σM, σV and σX monitored by the expression of lacZ placed under the control of the respective σ factor gene promoter and integrated into the chromosomal amyE locus decreased with increasing IPTG concentration  . A ugtP mutant gene whose 18th codon was changed from His to Ala produced a normal level of the protein but showed no glucolipid synthetic activity. When Pspac-ugtPH18A-His6 was integrated into the aprE locus instead of Pspac-ugtP-His6, the activities of σM, σV and σX did not change with increasing IPTG concentration. Thus, the lack of glucolipids, not the absence of UgtP protein, activates these ECF σ factors  .
Escherichia coli has a much simpler membrane lipid composition than B. subtilis. The membrane contains three kinds of phospholipids (phosphatidylethanolamine, phosphatidylglycerol and cardiolipin), except for the outer leaflet of the outer membrane, which is composed of lipopolysaccharide  . It does not contain glucolipid. When the genes for B. subtilis σM, σV and σX and their anti-σ factors were introduced into E. coli cells, lacZ fused to the ECF σ factor-regulated promoters was expressed. Additional introduction and expression of B. subtilis ugtP gene in the E. coli cells led to the synthesis of small amounts of glucolipids, and the activities of B. subtilis σM and σV were partially repressed, but the activity of σX was unaffected  . This result indicates that in heterologous E. coli cells the lack of glucolipids activates B. subtilis σM and σV.
We wondered if B. subtilis σV is activated by an RIP mechanism in E. coli cells. Although E. coli has a signal peptidase and a site-2 protease RseP, which functions in RIP of the anti-σ factor to ECF σ factor σE    , we thought it possible that the activation of σV might not involve RIP. In the same vein, we suspected that the activation of σV in a B. subtilis ∆ugtP mutant lacking glucolipid  might not involve RIP either. It seemed possible that a conformational change of anti-σV activates σV weakly; σV activation by the lack of glucolipids is only two- to three-fold, whereas treatment with a high concentration (100 µg/ mL) of lysozyme causes ~65-fold activation  . In this study we explored the question of how σV is activated in the B. subtilis ∆ugtP mutant.
2. Materials and Methods
2.1. Bacterial Strains, Plasmids and Culture Media
The strains and the plasmids used for this study and DNA primers used for strain construction are listed in Tables 1-3, respectively. The marker-free deletion of rsiV (SBS352) was obtained as previously described   and confirmed by polymerase chain reaction (PCR). Luria-Bertani (LB) medium was used. For plates, media were solidified with 1.5% agar. Tryptose blood agar (Difco) was used for preparation of B. subtilis strains for transformation. Synthetic media CI and CII were used for competence development for transformation  . When appropriate, antibiotics were included at the following concentrations (in µg/mL): ampicillin (100), chloramphenicol (5), erythromycin (0.5), neomycin (10), spectinomycin (100) and tetracycline (10). Lysozyme used was Wako Pure Chemical’s lysozyme hydrochloride from egg white. Cell growth was monitored with a mini photo 518R photometer (TAITEC) equipped with a 530- nm interference filter.
2.2. Construction of Chimeric Anti-σ Factors
RsiV and RsiW were divided into N-terminal regions (RsiV1-36 and RsiW1-82), transmembrane regions (RsiV37-59 and RsiW83-105) and C-terminal regions (RsiV60-285 and RsiW106-208), respectively, based on the prediction of the transmembrane regions by TMHMM (URL:
http://www.cbs.dtu.dk/services/TMHMM/). Chimeric anti-σ factors, a chimera of the N-terminal region of RsiV and the transmembrane and C-terminal regions of RsiW (RsiVN-RsiWTMC) and a chimera of the N-terminal and transmembrane regions of RsiV and the C-terminal region of RsiW (RsiVNTM-RsiWC), were designed. These chimeras were constructed by PCR and Gibson Assembly Cloning Kit (New England Biolabs) on a vector pSG1729  digested with KpnI and BamHI. The chimeric genes were transferred together with ribosome-binding site (RBS) on the vector pSG1729 by PCR and Gibson Assembly Cloning Kit to a vector pAPNC213  digested with SalI. The resultant chimeric genes under
Table 1. Bacterial strains used for this study.
a ®, transformation with chromosomal or plasmid DNA.
Pspac and RBS were integrated into the chromosomal aprE locus. Constructs were verified by sequencing.
2.3. Genetic and Recombinant DNA Procedures
These were based on standard methods   . The cycle number and the annealing temperatures for PCR were 40 cycles and 40˚C - 65˚C according to the Tm values of primers, respectively.
2.4. Biochemical Procedures
The β-galactosidase assay method using o-nitrophenyl-β-d-galactoside as sub-
Table 2. Plasmids used for this study.
Table 3. PCR primers used for strain construction.
strate and the unit definition were as described by Wang and Doi  . Proteins were separated by SDS gel electrophoresis with 15% polyacrylamide gel. Luminata Forte Western HRP Substrate (Millipore) was used for immunodetection according to the manufacturer’s instructions. For immunoblotting, anti-DYKDDDK tag (FLAG tag) monoclonal antibody (Wako Pure Chemicals) and anti-σA antiserum (a gift from K. Asai) were used. The secondary antibodies were peroxidase-conjugated anti-mouse IgG (H+L) and anti-rabbit IgG (H+L) (Jackson ImmunoResearch Laboratories). Immuno-stained bands were detected by C-Digit scanner (LI-COR), and the band intensities were measured by Image Studio software (LI-COR).
3.1. Lack of Glucolipids Affects σV Activity through the Anti-σ Factor RsiV
The activity of σV was monitored by β-galactosidase activity expressed from PsigV-lacZ integrated into the chromosomal amyE locus. A ugtP deletion mutant showed an approximately two-fold activation of σV compared to wild type (Figure 1(a)). In ∆sigV and ∆sigV ∆ugtP mutants the activity was almost abolished, indicating that PsigV-lacZ expression was dependent on σV activation. Deletion of the rsiV gene encoding anti-σV factor led to high σV activity, but additional introduction of a ugtP deletion mutation did not lead to additional activation of σV (Figure 1(a)). The rsiV deletion mutant was insensitive to the lack of glucolipids, which is consistent with the idea that the lack of glucolipids affects σV activity through RsiV.
3.2. 1,10-Phenanthroline Inhibited Activation of σV Caused by Addition of Lysozyme But Not the Activation by Lack of Glucolipids
The zinc chelator 1,10-phenanthroline (PT) inhibits the site-2 protease RasP  , as it does many other site-2 metalloproteases  . When ugtP+ cells were challenged by hen egg white lysozyme (HEWL) at low concentration (100 ng/mL), σV was activated several-fold. Addition of 10 mM PT returned the σV activity to the basal level (Figure 1(b)), most probably via inhibition of RasP. Approximately two-fold activation of σV in the ∆ugtP mutant compared to wild type was not changed by addition of 10 mM PT (Figure 1(b)), suggesting that the inhibition of RasP by PT did not affect activation. It seems that the activation of σV by the lack of glucolipids does not depend on the function of the site-2 protease RasP. Addition of HEWL to the PT-treated ∆ugtP cells did not increase the σV activity (Figure 1(b)). The σV activation by HEWL seems to depend on RasP in the ∆ugtP cells as well as in the ugtP+ cells.
3.3. A rasP Mutation Abolished the σV Activation Caused by the Addition of Lysozyme But Not the Activation by the Lack of Glucolipids
When wild-type cells were treated with increasing concentrations (25, 50 and
Figure 1. Activation levels of σV monitored via PsigV-lacZ transcriptional fusion. Cells were grown in LB medium to mid-exponential phase (mini photo 518R reading of 0.35 - 0.45) unless otherwise noted, and b-galactosidase activity was measured. The means and standard deviations of at least three measurements are shown. Asterisks indicate significant differences as determined by Student’s t-test: *P < 0.05, **P < 0.01. (a) Effect of ∆sigV and ∆rsiV on the sV activity. Strains YTB041, YTB040, SBS313, SBS314, SBS361 and SBS362 were analyzed; (b) Effect of PT on the sV activity. Strains YTB041 and YTB040 were analyzed. For HEWL treatment of YTB041 the cells were grown to early-exponential phase (mini photo reading of 0.20), and HEWL was added to 100 ng/mL, followed by 30 min shaking, after which the culture reached mid-exponential phase. For PT treatment of YTB040 the cells were grown to mid-exponential phase (mini photo reading of 0.45), and PT was added to 10 mM, followed by 20 min shaking. For HEWL and PT treatment the cells were grown to mid-exponential phase (mini photo reading of 0.45), and PT was added to 10 mM, which stopped the turbidity increase; after 20 min shaking HEWL was added to 100 ng/mL, followed by 30 min shaking; (c) Effect of a rasP mutation on the σV activity. Strains YTB041, SBS101, YTB040 and SBS102 were analyzed. The cells were treated with HEWL at the concentrations (in ng/mL) indicated in the square brackets as described for YTB041 in (b); (d) FLAG-RsiV is functional. Strains SBS204 and SBS205 were analyzed. The cells were grown in the presence or absence of 1 mM IPTG; (e) Anti-σ activities of RsiVN-RsiWTMC and RsiVNTM-RsiWC. Strains SBS500, SBS501, SBS502 and SBS503 were analyzed. One millimolar IPTG was added or withheld from the beginning of the culture. SBS500 and SBS502 cells were subjected to 100 ng/mL HEWL as described for YTB041 in (b). For alkali treatment of SBS500 and SBS502 the cells were grown to early-exponential phase (mini photo reading of 0.20) and 5.5 µl of 5 N NaOH was added to the 5-mL cultures, which brought the medium pH to ca. 8.9; this was followed by 30 min shaking.
100 ng/mL) of HEWL, the activity of σV increased accordingly (Figure 1(c)). A rasP::tet mutant showed a basal level σV activity in the absence of HEWL (Figure 1(c)). Whereas the addition of 100 ng/mL HEWL activated σV several-fold in rasP+ cells, the σV activity level in the rasP::tet mutant did not change in the presence of 100 ng/mL HEWL (Figure 1(c)), indicating that the activation of σV by HEWL requires RasP  . On the other hand, the approximately two-fold activation of σV in the ∆ugtP mutant did not change when, in addition, rasP was disrupted (Figure 1(c)), indicating that RasP has nothing to do with the σV activation by the lack of glucolipids. Addition of 100 ng/mL HEWL to the ∆ugtP mutant resulted in a remarkably high activation of σV (Figure 1(c)). The ∆ugtP mutant seems to become sensitized to the stress from HEWL addition. Addition of HEWL to ∆ugtP rasP::tet mutant cells did not change the σV activity (Figure 1(c)). The σV activation by HEWL required RasP in the ∆ugtP cells as well as in the ugtP+ cells.
3.4. RsiV Was Not Degraded in the ∆ugtP Cells
RsiV is degraded by RIP in response to lysozyme. N-terminally FLAG tag-fused RsiV was expressed and analyzed by immunoblotting (Figure 2). When ugtP+ cells were treated with 25, 50 and 100 ng/mL HEWL, the amount of intact FLAG-RsiV protein was reduced to about 90%, about 70% and about 40%, respectively, compared with cells that were not treated with HEWL. This is consistent with the increasing activity of σV with increasing concentrations of HEWL (Figure 1(c)). In contrast, ∆ugtP cells showed no decrease in the amount of intact FLAG-tagged RsiV compared with the wild type not treated with HEWL. This result indicates that site-1 cleavage did not occur in ∆ugtP cells. Site-1 cleavage is a prerequisite for site-2 cleavage, and RasP cleavage is constitutive once site-1 cleavage has occurred  . The idea that RasP has nothing to do with the σV activation by lack of glucolipids fits into this scenario. When 100 ng/mL HEWL was added to ∆ugtP cells, the amount of intact FLAG-RsiV protein was reduced to about 7% (Figure 2). This corresponds well with the remarkably high activation of σV in the ∆ugtP mutant when 100 ng/mL HEWL was added (Figure 1(c)). The FLAG-tagged RsiV was functional: when the expression of Pspac-flag-rsiV was induced with IPTG, σV activity was repressed in both ugtP+ and ∆ugtP cells (Figure 1(d)).
Figure 2. Degradation of RsiV. Strains SBS200 and SBS201 were analyzed. The cells were grown in LB medium containing 1 mM IPTG to mid-exponential phase (mini photo reading of 0.40 - 0.45), and 250 µg/mL chloramphenicol and HEWL of the indicated concentrations were added, followed by 10 min shaking. Harvested samples were sonicated and subjected to SDS-PAGE. The sample amount per lane was normalized by the culture turbidity. FLAG-RsiV and σA (an internal control) were detected by immunoblotting. The relative ratio of the amount of Flag-RsiV to that of σA is shown under each lane. The ratio in SBS200 (WT) in the absence of HEWL was set to 1.0.
3.5. The Behavior of RsiV-RsiW Chimeric Anti-σ Factors Indicated That the C-Terminal Extracytoplasmic Region of RsiV Was Necessary for the Response to Lack of Glucolipids
Chimeric anti-σ factors, RsiVN-RsiWTMC and RsiVNTM-RsiWC, were constructed (see Materials and Methods). RsiW is an anti-σ factor for σW   . Whereas σW is activated by RIP of RsiW   in response to a variety of signals, including an alkaline shock  , it is not activated by lack of glucolipids   .
Both chimeric anti-σ factors are functional as anti-σV: when their expression was induced with IPTG in a ∆rsiV background, σV activity was repressed (Figure 1(e)). The N-terminal region of RsiV seems to be sufficient to sequester σV.
In cells expressing either chimeric anti-σ factor, σV was not activated by the addition of 100 ng/mL HEWL (Figure 1(e)). This is consistent with reports that the C-terminal region of RsiV binds HEWL   .
We found that σV was activated when the cells expressing RsiVN-RsiWTMC suffered alkaline stress (pH 8.9) (Figure 1(e)). Most probably the transmembrane and C-terminal extracytoplasmic regions of σW underwent RIP in response to the alkaline stress, which led to degradation of the chimeric anti-σ factor and to release of σV. In contrast, in the cells expressing RsiVNTM-RsiWC σV was not activated in response to alkaline stress (Figure 1(e)). The C-terminal extracytoplasmic region of RsiW may undergo site-1 cleavage, but it does not seem to lead to the site-2 cleavage of the fused transmembrane region of RsiV.
The expression of the chimeric anti-σ factors in ∆ugtP cells resulted in σV activity at the basal level of the ugtP+ cells (Figure 1(e)). This indicates that the C-terminal extracytoplasmic region is necessary for σV to respond to the lack of glucolipids.
We also constructed chimeric anti-σ factors, RsiWN-RsiVTMC and RsiWNTM- RsiVC, but they were not functional as anti-σW: IPTG induction of their expression in a ∆rsiW background even from multicopy plasmids did not lead to repression of σW activity (data not shown); the reason is unknown.
B. subtilis σV is activated primarily in response to lysozyme  . The activation depends on the RIP of the anti-σ factor RsiV    . By contrast, in a ∆ugtP mutant that lacks glucolipids, σV is activated without RIP of RsiV. The activation was not inhibited by the zinc chelator PT, an inhibitor of the site-2 protease RasP, nor was it abolished in a rasP mutant. No proteolysis of RsiV was observed in the ∆ugtP mutant, indicating that site-1 cleavage did not occur. Most probably a conformational change of RsiV without proteolysis is responsible for the release of σV from RsiV in the ∆ugtP mutant.
Some relevant information concerning the activation of B. subtilis ECF σ factors is available. It is known that σW is activated by RIP. Under alkaline stress, the anti-σ factor RsiW is first degraded by the site-1 protease PrsW, trimmed by several peptidases, and then degraded by the site-2 protease RasP  . There are two anti-σ factors, YhdL and YhdK, involved in the regulation of σM, where YhdL has the major function of binding σM directly. Under salt stress, σM is activated but no proteolysis of N-terminally hexahistidine-tagged YhdL is observed  . This suggests that σM is probably released by a conformational change of YhdL.
The activation of σV in the ∆ugtP mutant cells upon addition of 100 ng/mL HEWL was higher than in ugtP+ cells, and degradation of RsiV was correspondingly enhanced in the ∆ugtP mutant cells. This may be due to the proposed conformational change of RsiV in the ∆ugtP mutant cells.
Addition of Mg2+ suppressed the σV activation in the absence of glucolipids  . Thus, it seems unlikely that direct interaction of glucolipids with RsiV is required for its sequestering of σV. The lack of glucolipids probably affects the membrane properties, which may then lead to a conformational change of RsiV and to release of σV. One line of reasoning might be as follows: glucolipids have no charge in their polar head groups. It is possible that the change in the charge distribution on the membrane surface brought about by their absence affects the conformation of RsiV. We note that synthesis of Acholeplasma laidlawii monoglucosyldiacylglycerol suppressed, albeit only partially, the σV activation in the ∆ugtP mutant  . This too, may be caused by a change in the charge characteristics of the membrane surface. In this regard, it is also noteworthy that the membrane surface charge characteristics are an important topological determinant of E. coli LacY protein  .
Experiments with chimeric anti-σ factors indicated that the C-terminal region of RsiV is necessary for the response to glucolipid deficiency. It seems that the C-terminal extracytoplasmic region of RsiV is important for it to be sensitive to the alteration in charge characteristics of the membrane surface.
We conclude that, in glucolipid-lacking mutants of B. subtilis, σV can be activated without proteolysis of anti-σ factor RsiV and that this is most likely brought about by a conformational change of RsiV due to the alteration of the membrane surface charge characteristics.
We are grateful to Thomas Wiegert for providing the chromosomal DNA of the strain 1012 rasP and to Kei Asai for providing the strain SBS200 and anti-σA antiserum. T. S. is a research fellow of Japan Society for the Promotion of Science (JSPS). This work was supported in part by JSPS KAKENHI (Grant-in-Aid for JSPS Research Fellow) Grant Number JP16J10629 to T. S.