Bacillus subtilis cells grown under phosphate starvation induce teichuronic acid (TUA) synthesis while simultaneously repressing teichoic acid synthesis (TA). The turnover rates of TA-containing and TUA-containing walls are similar, indicating that autolysin function is similar and suggesting that modulation of autolytic function may be similar. In this study, it is demonstrated, utilizing fluorescein isothiocyanate (FITC)–dextran to probe the wall pH, that a low pH exists in the wall matrix. A second probe, cationized ferritin (CF), was used to observe cell surface protonation. Suspensions of B. subtilis cells containing either TA or TUA were aggregated with CF only after the addition of a protonmotive force-dissipating agent. Respiring B. subtilis TUA-containing cells labelled with FITC–dextran exhibited little fluorescence. Conversely, fluorescence intensities exhibited by cells de-energized with nitrogen gas were significantly greater. The effects of protonmotive force on autolytic activity were studied by adding cell wall protein extract containing concentrated autolysin to exponentially growing TA-containing and TUA-containing B. subtilis cells. Both TUA-containing and TA-containing cells were lysed only after the addition of sodium azide. These data suggest that during normal growth the wall of TUA-containing B. subtilis cells is protonated, and protonmotive force influences autolytic regulation in both TUA-containing and TA-containing B. subtilis cells.
Growth of Bacillus subtilis in phosphate-limited medium results in cells with walls containing teichuronic acid (TUA) instead of teichoic acid (TA). TUA synthesis begins at the same time as phosphate limitation, while the TA-containing portion of the wall is excised due to turnover (Ellwood and Tempest, 1969). Cell walls containing TUA exhibit turnover rates similar to those containing TA. It is thus clear that autolysins of TUA-containing walls function in the manner similar to the autolysins in TA-containing walls (Mauck and Glaser, 1970). Nothing is known about the control of autolysin when bacilli are grown in phosphate-restricted media. Recently, Chambert and Petit-Glatron (1999) provided evidence that teichoic acid assists in the folding of secreted enzymes, suggesting that the folding of autolysins on the outside of the Bacillus subtilis cell may be influenced by anionic polymers such as teichoic acids. More recently, it has been shown that D-alanylation of teichoic acids enhances the rate of folding of secreted proteins (Hyyrylainen et al., 2000). In 1981, Jolliffe et al. reported that any condition leading to dissipation of the protonmotive force (PMF) induced autolysis in B. subtilis. Later, Kemper et al. (1993) suggested that the wall was a reservoir for protons secreted during normal respiration, giving rise to walls with a pH lower than that of the surrounding growth medium. As a test of this hypothesis, Urrutia et al. (1992) showed that respiring B. subtilis had a much lower affinity for uranyl ion than cells with no PMF. Hydrogen ion-sensitive chemical and fluorescent probes were recently employed by Calamita et al. (2002) to show that the cell walls of exponentially growing B. subtilis were protonated. The results of Calamita et al. (2002) are now extended to show for the first time a role for PMF in the regulation of autolysins of Bacillus subtilis grown under conditions of phosphate limitation. It is now shown that autolysins in TA-deficient, TUA-sufficient walls are also influenced by PMF.
FACS analyses of de-energized B. subtilis cells
FITC, a pH-sensitive dye, conjugated to dextran with the molecular weight of 10 000 Da was used to probe the cell wall environment of B. subtilis cells containing TUA. Demchick and Koch (1996) determined that the rate of penetration of dextrans into the cell wall matrix of B. subtilis is a function of dextran molecular weight. FITC–dextran does not penetrate the cell but does become distributed in the cell wall matrix. Thus, FITC–dextran is a reporter for hydrogen ions. FITC-saturated walls acidified by secreted protons (respiring cells) would be expected to have very little, if any, fluorescence. In contrast, FITC-saturated walls from non-living cells would be expected to fluoresce because these walls are no longer acidic and would possess a pH identical to that of their surrounding medium. The walls of B. subtilis cells grown in phosphate-limited medium were loaded with FITC–dextran and de-energized with nitrogen gas and the resulting fluorescence intensity of the cells was measured using FACS. Cells with de-energized membrane exhibited markedly greater fluorescence intensities than control cells (Fig. 1A). In separate control experiments, FACS analyses (Calamita et al. 2002) of B. subtilis cells grown in Penassay broth revealed results similar to those seen with the B. subtilis cells grown in phosphate-limited medium. PMF was restored with phenazine methosulphate and glucose in de-energized cells. Upon restoration of PMF the fluorescence intensities decreased significantly (Fig. 1B).
Aggregation of cells with cationized ferritin
Suspensions of B. subtilis cells grown in either phosphate-sufficient or phosphate-limited medium were mixed with cationized ferritin (CF). Within several minutes of the addition of sodium azide to cell suspensions, aggregates formed, whereas control cells remained suspended. Similar results were observed for cells grown in phosphate-sufficient or phosphate-limited medium, independent of their wall composition (Fig. 2). Thus, cationized ferritin-promoted aggregation is not a function of wall composition but depends on the presence of negative charges in the cell wall. The loss in opacity was not due to cellular lysis, as lysis was less than 10% under conditions of the experiment as determined by turbidity measurements. Control experiments using exponential-phase cells grown in the presence of [14C]-proline did not release 5% trichloroacetic acid-insoluble protein into the buffer. Furthermore, upon light microscopy there was evidence of massive aggregation of B. subtilis cells.
Induction of lysis with addition of concentrated autolysin to B. subtilis cells containing either TA or TUA
Partially purified autolysin was extracted from isolated cell walls and adjusted to a protein concentration equivalent to 100 times higher than that found in the walls of late-exponentially growing cells. This highly concentrated form of autolysin was active on cell wall preparations. Control experiments utilizing sodium dodecyl sulphate (SDS)-extracted walls revealed that the concentrated autolysin rapidly solubilized the walls of cells grown in phosphate-sufficient and phosphate-limited medium. Late-exponentially growing cultures of B. subtilis cells were treated with sodium azide, concentrated autolysin plus sodium azide, or concentrated autolysin only. Cultures treated only with the concentrated autolysin exhibited growth patterns virtually identical to those of control cells. Cultures with concentrated autolysin exhibited lysis only when an uncoupling agent such as sodium azide was added to the culture. Cultures with an uncoupling agent plus concentrated autolysin exhibited a more rapid rate of lysis than did cultures with uncoupler only. Results were similar between cultures grown in Penassay broth or phosphate-limited medium (Fig. 3). The lysis rate for cells grown in phosphate-sufficient medium was 4.7 × 10–3 min–1 in the absence of concentrated autolysin and was 1.03 × 10–2 min–1 in the presence of concentrated autolysin with the addition of sodium azide. Similarly, the rate of lysis following addition of sodium azide to cells grown in phosphate-limited medium was 7.7 × 10–3 min–1, whereas cell suspensions with added autolysin lysed at a rate of 2.76 × 10–2 min–1. In independent experiments, mid-exponential cultures were used, giving similar results. We preferred to use late-exponential phase cells because maximal autolysin activity is observed.
Chemical composition of walls obtained from phosphate-sufficient and phosphate-limited media
Walls from cells grown in phosphate-sufficient medium were readily aggregated by 25 μg ml–1 concanavalin A. In contrast, 500 μg ml–1 concanavalin A did not aggregate walls obtained from cells grown in phosphate-limited medium. Table 1 shows the phosphorus, hexose and uronic acid compositions from B. subtilis 168 cells grown in Penassay broth or minimal medium with limited phosphate. In TA-containing walls there was 1.56 μmol of hexose per milligram of cell wall, 1.48 μmol of phosphorus per milligram of cell wall and no uronic acid. The TUA wall was rich in uronic acid but contained no hexose. These walls, however, contained only 0.4 μmol of phosphorus per milligram of cell wall. Similar results have been reported by Mauck and Glaser (1972).
Table 1. Composition of the cell walls of Bacillus subtilis 168a.
Bacillus subtilis 168 cells were grown in either phosphate-sufficient or phosphate-restricted minimal medium. Cells were harvested, washed twice with distilled water and broken with a French pressure chamber. Walls were obtained by differential centrifugation and extracted in boiling 0.1% SDS. The insoluble walls were then extracted several times in hot water and freeze-dried. The numbers shown above are in μmol per mg of wall.
In B. subtilis, cell cylinder elongation occurs by inside-to-outside growth, concomitant with the turnover of stress-bearing older wall on the cell periphery (Koch and Doyle, 1985). In order for the bacteria to maintain cell wall integrity during elongation, it is important that the autolytic enzymes are well regulated in both TA- and TUA-containing walls. It has been suggested that one factor that may regulate autolysin is a local low pH in respiring cells. Secreted proteases are also known to reduce autolysin levels. These proteases, however, act at the very periphery of the cells, and reduce cell wall turnover (Jolliffe et al., 1980; Stephenson et al., 1999). In respiring cells, protons are extruded across the cytoplasmic membrane and bind to cell wall constituents (Jolliffe et al., 1981; Kemper et al., 1993; Calamita et al., 2002). In TA-containing cell walls protons would be expected to bind to phosphate and any free carboxylate of peptidoglycan. In TUA-containing cell walls, protons would be expected to bind to uronic acids as well as carboxylates from peptidoglycan. In either case, protonation of either TUA- or TA-containing cell wall would be expected to create a wall matrix of low pH. Protonmotive force sustains this local low pH and, in turn, autolytic activity is inhibited because N-acetylmuramyl-L-alanine amidase cannot function below pH 5.5. Chambert and Petit-Glatron (1999) provided evidence that teichoic acid facilitated the folding of secreted enzymes as the enzymes traversed the wall. Because autolysins are also secreted and act at the very periphery of cells, it is possible that teichoic acid may be involved in the normal folding of autolytic enzymes. Control experiments established that B. subtilis containing at least 100 times normal concentration of autolysin grew at the same rate as cells without added autolysin. Loss of PMF induced by sodium azide (Fig. 3) resulted in a more rapid rate of autolysis in the presence of exogenously added autolysin. Thus, even exogenously added concentrated autolysin appears to be under the influence of PMF.
By using a pH-sensitive probe (FITC–dextran), it was shown that the pH of the cell wall in cells grown under conditions of phosphate limitation is probably low. This is supported by the fact that dissipation of PMF by nitrogen gas resulted in a marked increase in fluorescence of the probe (Fig. 1). As further evidence of the role of PMF in creating an environment around the cell with a relatively low pH, suspensions re-energized with phenazine methosulphate exhibited less fluorescence in the presence of FITC–dextran (Fig. 1).
Another probe for cell0surface pH, cationized ferritin, was employed in cultures grown in phosphate-sufficient and phosphate-limited media. If the cell surface is protonated, then very little of the CF would be expected to bind to B. subtilis. On the other hand, cells unable to maintain PMF should have the same cell-surface pH as the surrounding medium. The results showed that CF readily aggregated both TA- and TUA-containing cells upon the addition of sodium azide (Fig. 2). Cells undergoing normal respiration tended to be refractory to aggregation by CF.
Table 1 shows that cells grown under phosphate-limitation contained some phosphate in their walls. The origin of the phosphate is uncertain, but it is likely that the phosphate originates from linkages between muramic acid and teichuronic acid (Lang et al., 1982). The inability of some strains of B. subtilis 168 to completely shut down teichoic acid synthesis during conditions of phosphate limitation is also likely to contribute to the phosphate detected in the wall (Lahooti and Harwood, 1999). Furthermore, cells grown under phosphate limitation did not aggregate with concanavalin A, suggesting the absence of α-glucosylated teichoic acid (Doyle and Birdsell, 1972).
The composite interpretation of the results of this report is that cells with a normal PMF are able to regulate their autolysins. Loss of PMF results in cellular lysis. It is suggested that during metabolism cells secrete protons, which bind to anionic sites in the cell wall. As long as cells are able to maintain PMF the walls remain in a protonated state. The results do not exclude any role of anionic polymers in the wall in assisting the folding of secreted proteins, including autolysin, in B. subtilis (Chambert and Petit-Glatron, 1999; Jolliffe et al., 1981). These are the first results suggesting that autolysins are regulated by protonmotive force in both TUA- and TA-containing cell walls.
Bacterial strains and chemicals
Bacillus subtilis strains 168 (trpC2) and SR 22 (trpC2, spoOA12) were obtained from the Bacillus Genetic Stock Center, Ohio State University, and maintained by regular transfer on Penassay agar. All salts were of reagent grade. Fluorescein isothiocyanate-labelled dextran (MW 10 000) was purchased from Sigma Chemical Co. (St. Louis, MO, USA)
Cells were grown in minimal medium with limiting phosphate (Mauk and Glaser, 1970) containing glycerol (6 g l–1) and supplemented with tryptophan (50 μg ml–1). Cultures grown in minimal medium with limiting phosphate were transferred daily for a minimum of 4 days prior to use in experiments.
Labelling cells with FITC–dextran
Bacillus subtilis 168 cells were grown overnight in phosphate-limited minimal medium. Prewarmed medium (20 ml) containing 36 mg of filter-sterilized FITC–dextran was inoculated with 0.5 ml overnight culture. The cells were grown to exponential phase at 37°C in a gyratory shaker (200 r.p.m.).
De-energization studies with FITC–dextran-labelled cells
Two 5-ml culture samples of cells in the medium were placed in two glass tubes (13 × 100 mm) and maintained on ice. One tube was covered with Parafilm, and nitrogen gas was introduced to the cells through the Parafilm using a Pasteur pipette. For examination by flow cytometry, 1 ml of culture was placed in a microfuge, washed and suspended in 500 μl of PBS, pH 7.3 (50 mM phosphate, 150 mM NaCl), containing 1.8 mg ml–1 unlabelled dextran (MW 10 000) while all the time preventing the introduction of oxygen. The cells were immediately subjected to FACS analyses (Becton-Dickinson FACScan with a 15 mW argon laser tuned to 488 nm). To re-energize the cells, suspensions were washed once and suspended in phosphate-buffered saline (PBS) containing 2 mg ml–1 of glucose, 1.8 mg ml–1 unlabelled dextran and 400 μM phenazine methosulphate (Jolliffe et al., 1981).
Aggregation of B. subtilis cells in the presence of cationized ferritin
Bacillus subtilis 168 was grown overnight in both phosphate-sufficient and phosphate-limited medium. Prewarmed medium (50 ml) was inoculated with 0.5 ml of overnight culture and cells were harvested when they reached mid-exponential phase. The cells were washed twice and suspended to an absorbance of 0.8 at 540 nm in PBS containing 2 mg ml–1 glucose (pH 7.3). Suspensions of cells in PBS were treated with cationized ferritin (100 μg ml–1) and/or sodium azide (40 mM). Aggregation was measured using a spectrophotometer at 540 nm.
Chemical analyses of cell wall components
Bacillus subtilis 168 cells were grown in either phosphate-sufficient or phosphate-limited minimal media. Cells were harvested, washed twice with distilled water and broken with a French pressure chamber. Walls were obtained by differential centrifugation and extracted in boiling 0.1% SDS. The insoluble walls were then extracted several times in hot water and freeze-dried. The walls were then subjected to phosphorus (Chen et al., 1956), hexose (Dubois et al., 1956) and uronic acid analyses (Blumenkrantz and Asboe-Hansen, 1973).
Induction of lysis with B. subtilis SR22 cells in phosphate-limited medium and Penassay broth
Bacillus subtilis SR22 cells were grown overnight at 37°C with shaking in either phosphate-limited minimal medium or Penassay broth. Four nephelometer flasks containing 8 ml of either phosphate-limited minimal medium or Penassay broth were prewarmed to 37°C. The flasks were inoculated with 0.8 ml of overnight culture. Cultures were grown to late-exponential phase were treated as follows: sodium azide only, sodium azide plus concentrated autolysin, concentrated autolysin only, and one flask was left untreated. The final concentration of sodium azide was 40 mM. The cell turbidity was measured every 30 min with a Klett photometer.
Extraction of autolysin from B. subtilis SR22
Autolysins were extracted from the cell wall of B. subtilis as described by Brown (1973). Briefly, the cells from three litres of B. subtilis SR22 (in Penassay broth) were harvested and supernatants discarded. Cells were washed once in PBS (pH 7.3), the pellet was suspended in 10 ml of 5 M LiCl and incubated at 4°C for 1 h. The cells were centrifuged at 14 000 g for 15 min The cells were discarded and the supernatant placed in dialysis tubing with a molecular weight cut-off of 12 000. The supernatant was dialysed against 2 L of 0.1 M NaCl for 12 h, changing the NaCl at the 6 h point. The amount of protein in the extract was determined by BCA protein analysis using crystalline bovine serum albumin (BSA) as a standard. It should be kept in mind that this protein extract contains not only autolysin but other surface extractable proteins as well.
The authors thank Dr Sam Wellhausen for performing the FACS analyses, and Professor Arthur Koch for discussions. The work was supported in part by the R.J.D. Research Fund.