INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Calmodulin, a heat-stable, acidic, Ca²⁺-binding protein has been found in a wide variety of eukaryotic organisms (23, 31, 32). This 148-amino-acid monomeric protein is involved in the activation of more than 20 enzymes which mediate a wide variety of metabolic processes (11, 23, 32, 53). The monomer is encoded in most organisms by a single gene which is highly conserved throughout evolution (31), including the lower eukaryotes (9). Disruption of a calmodulin gene has proved lethal in three genera of fungi (13, 42, 49). It appears that calmodulin plays a central role in the regulation of the cell cycle and nuclear division (1, 27, 37).

Evidence for the presence of calmodulins in prokaryotic cells has been increasing in recent years (3). Onek and Smith (38) thoroughly reviewed the earlier evidence for the existence of calmodulinlike proteins in seven genera of bacteria. In the last 8 years, further evidence for calmodulinlike proteins has appeared in Mycobacterium smegmatis (7), Mycobacterium tuberculosis (16, 17), and Mycobacterium phlei (46); in Escherichia coli which has been induced with EGTA (26); in Nostoc sp. strain PCC 6720 (39); in three species of Bordetella, including B. pertussis (33, 34); and in Halobacterium salinarium (44). The potential roles of bacterial calcium-binding proteins, including calmodulins, were briefly reviewed recently (35).

Calcium ions play an important role in the metabolism of Bacillus subtilis cells. While calcium ions do not appear to be necessary for vegetative growth, they are essential for efficient protein degradation during sporulation and in the formation of heat-resistant spores (36). Furthermore, Ordal demonstrated earlier (40) that calcium ions induce a negative chemotactic response in vegetative cells of B. subtilis. Such observations led us to search for a calmodulinlike activity in this organism. A calmodulinlike protein was isolated from sporulating B. subtilis cells and was shown to have a molecular mass of 23 kDa, to stimulate phosphodiesterase from bovine brain and NAD kinase from pea in a dose- and Ca²⁺-dependent manner, to have a pI of 4.9 to 5.0, and to cross-react with bovine brain anticalmodulin antibodies (19). A heat-stable, hydrophobic, Ca²⁺-binding protein having an apparent molecular mass of 24 kDa has been isolated from the spores of Bacillus cereus (47); this calmodulinlike protein has been proposed to play a role in the release of calcium ions during spore germination (48).

In addition, a series of studies have demonstrated that B. subtilis cells have electrogenic and other calcium pumps (14, 15, 24, 29). It has been proposed (43) that during vegetative growth, the predominant function of a Ca²⁺-H⁺ antiport system is to maintain a low concentration of calcium ions in the cytosol, whereas during sporulation, a Ca²⁺ uniporter attains dominance and causes accumulation of calcium ions in the cytosol. In this study, we present immunocytochemical evidence that the calmodulin of B. subtilis cells (BsCaM) is produced constitutively during growth and sporulation and that it is localized in or near the cell membrane of late-log vegetative cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strain and culture methods. A stock culture of B. subtilis 3036 (trpC₂ isp1) was kindly provided to us by Mark Ruppen of Genetech, Inc. (2) and was used to avoid possible degradation of BsCaM by the predominant intracellular serine proteinase, ISP-1. Cells were cultured in chemically defined sporulation medium, CDSM (21), to eliminate any possible introduction of calmodulin from eukaryotic sources (18). B. subtilis 3036 cells were grown in 2-liter, triple-baffled flasks at 220 rpm at 37°C in a New Brunswick G-25 environmental shaker. When culture turbidity reached an A₆₅₀ of 1.4 to 1.6, cells were harvested by centrifugation (3,000 × g for 10 to 15 min) at 4°C. Cell pellets were resuspended in 2 M KCl and centrifuged as before; cell pellets were stored at 20°C. Thawed cells were used for protein purification by protocol I or II as described below.

Crude extract preparation and partial purification of the calmodulinlike protein from B. subtilis 3036. (i) Protocol I. Frozen pellets were resuspended in lysis buffer (2 M urea, 60 mM 2-mercaptoethanol, 1 mM CaCl₂, and 2 mM phenylmethylsulfonyl fluoride [PMSF]), stirred for 4 h at room temperature, and centrifuged at 35,000 × g at 4°C for 1 h. The supernatant fraction was dialyzed against three changes of a mixture of 20 mM Tris, 2 mM CaCl₂, and 2 mM PMSF (pH 7.5), overnight. The dialyzed sample was heated for 10 min in an 85°C water bath, immediately placed on ice for 30 min, and centrifuged at 35,000 × g for 30 min at 4°C. The supernatant fraction was brought to 70% (wt/vol) saturation with (NH₄)₂SO₄ at 0°C, and the resulting precipitate was removed by centrifugation at 35,000 × g for 1 h at 4°C. Salt was removed from the supernatant solution by dialysis against a mixture of 10 mM Tris-HCl, 2 mM PMSF, and 1 mM CaCl₂ (pH 7.5). After dialysis, the pH of the dialyzed sample was adjusted to 4.3 by the addition of 1 M acetic acid and stirred on an ice bath for 45 to 60 min. The sample was centrifuged at 35,000 × g at 4°C, and both the acid pellet and supernatant fractions were retained. The pH of both fractions was readjusted to 7.5. The acid supernatant fraction was loaded into a Blue 72 affinity chromatography column (Sigma) and eluted with 10 mM Tris, 1 mM CaCl₂, and 2 M urea. The eluted sample was dialyzed against a mixture of 20 mM Tris, 2 mM CaCl₂, and 2 mM PMSF to remove urea. Both the affinity-purified material and crude extracts were reacted against anticalmodulin antibodies and assayed for Ca²⁺ binding and phosphodiesterase activity.

(ii) Protocol II. Cells were grown, harvested, and lysed as described for protocol I. The heating and ammonium sulfate precipitation steps were omitted. The supernatant fraction was adjusted to pH 4.3 immediately after lysis and stirred at 0°C for 1 h. The precipitated solution was centrifuged at 35,000 × g for 1 h, and both supernatant and pellet fractions were collected. Part of the sample was loaded onto a Blue 72 affinity chromatography column (Sigma Chemical Co.), and the rest of the sample was kept at 20°C for further testing.

Immunological methods. (i) Dot blotting. Samples were blotted onto nitrocellulose membrane BA-S83 (Schleicher & Schuell). The nitrocellulose was incubated in 0.2% glutaraldehyde for 15 to 20 min at room temperature. After glutaraldehyde fixation, the membrane was briefly rinsed in 20 mM Tris-500 mM NaCl (pH 7.5) (TBS) buffer. The nitrocellulose membrane was then reacted with primary antibody and bovine brain or bovine testis anticalmodulin antibodies (1:100 dilution) for 2 h at room temperature. Excess antibody was removed by washing with TBS several times. Incubation with secondary antibody conjugated to alkaline phosphatase was done for 1 h at room temperature. After several washes, the reaction was developed by the addition of 5-bromo-4-chloro-3-indolylphosphate (BCIP)-nitroblue tetrazolium (22).

(ii) ELISAs. Enzyme-linked immunosorbent assays (ELISAs) were carried out in Immulon IV 96-well microtiter plates. Samples were tested against bovine brain and bovine testis anticalmodulin antibodies. Microtiter plates were coated with 60 µl of each sample and incubated overnight at 4°C. Following incubation, wells were washed with TBS to remove excess protein. Unreacted sites in the wells were blocked by adding 300 µl of a solution containing 3% bovine serum albumin (BSA) (globin free) and 1% gelatin incubating for 3 h at room temperature. After incubation, wells were washed three times with TBS. Primary antibody (polyclonal anticalmodulin from bovine brain developed in goat) was added in a 1:100 dilution and incubated for 2 h at room temperature. To remove excess antibody, wells were washed with TBS three times, and secondary antibody (anti-goat immunoglobulin G [IgG] conjugated to alkaline phosphatase) was diluted 1:1,000 and incubated in the wells for 1 to 2 h at room temperature. Paranitrophenylphosphate (PNPP) was added as a substrate and incubated for 30 min at room temperature (22). The plates were read in a Bio-Tech plate reader at 405 nm.

(iii) Western blot assay. Samples were subjected to electrophoresis on discontinuous acrylamide-sodium dodecyl sulfate (SDS) microslab gels, using a 4% acrylamide stacking gel and a 15% acrylamide resolving gel according to the method of Laemmli (25). Western blotting was done according to the procedures of Van Eldik and Wolchok (51) with some modifications. Nitrocellulose membranes BA-S85 and BA-S83 (Schleicher & Schuell) were used. Transfer of proteins was conducted at 150 mA for 30 min at room temperature. The protein transfer was verified by staining the nitrocellulose membrane with Ponceau S. Incubation with primary antibody (1:100 dilution) was done overnight at 4°C. Secondary antibody incubation (1:1,000 dilution) was carried out for 2 h at room temperature. The enzyme-linked antibody was detected with BCIP-nitroblue tetrazolium as described above.

Protein detection on blotting membranes. Ponceau S staining is a reversible staining method that was used to verify transfer efficiency before proceeding with immunodetection. Staining procedures were performed by the method of Salinovich and Montelaro (45). Amido black (Sigma Chemical Co.) was used to verify the positions of bands after autoradiography. The same nitrocellulose membrane exposed to the X-ray film was stained with amido black to locate the radioactive bands. The nitrocellulose blot was immersed in amido black stain (0.1% amido black 10-B, 45% methanol, 10% acetic acid). The blot was incubated for 5 min at room temperature with agitation. Following incubation the nitrocellulose membrane was destained in a solution of methanol-acetic acid-water (90:2:8 [vol/vol/vol]), rinsed with water, and dried.

Electron microscopy. Vegetative or sporulating cells were fixed immediately upon harvesting in 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 h at 4°C. Samples were dehydrated in a series of graded ethanol solutions and embedded in LR white (Electron Microscopy Sciences). The plastic was incubated at 50°C for 24 h in gelatin capsules. Thin sections were cut with a diamond knife and mounted on nickel grids (Electron Microscopy Sciences) coated with nitrocellulose. Postembedding immunocytochemistry was performed by using these thin sections from LR white blocks. The sections were incubated for 2 h on a drop of primary antibody (a polyclonal preparation against calmodulin from bovine brain, developed in a goat) diluted 1:20 and 1:40 in TBS containing 0.05% Tween 20, 0.05% gelatin, and 1% BSA. The incubation was carried out in a moist chamber at room temperature. The sections were washed in TBS containing 0.05% Tween 20, 0.05% gelatin, and 1% BSA. After washing, the sections were incubated on a drop of secondary antibody (anti-goat IgG conjugated to 5-nm gold; Sigma Chemical Co.) in TBS for 50 min at room temperature. In a control experiment, the primary antibody was replaced with normal goat serum. Micrographs were taken in a Hitachi H 7000 at an acceleration voltage of 80 kV. Most of the immunocytochemistry was carried out at the New Mexico State University Electron Microscopy Facility.

Membrane vesicle preparation. B. subtilis cells were grown in CDSM at 37°C with vigorous aeration as described above. Cells were harvested at an A₆₅₀ of 1.0 to 1.5, washed with 0.1 M potassium phosphate buffer (pH 7.3), and centrifuged at 14,000 × g for 20 min at 4°C. Cells were resuspended in 0.1 M potassium phosphate buffer (pH 8.0) containing 20% sucrose at a concentration of 1.00 g (wet weight) per 80 ml. Lysozyme (Sigma Chemical Co.) was added to a final concentration of 300 µg/ml. The mixture was incubated for 30 min at 37°C. Protoplast formation was monitored by light microscopy. When the protoplast population was 90%, these were harvested by centrifugation at 16,000 × g for 30 min at 4°C. Protoplasts were resuspended in 40 ml of 50 mM potassium-phosphate buffer (pH 6.6). Lysed and whole cells were then removed by centrifugation at 900 × g for 30 min at 4°C. Approximately three-quarters of the supernatant fraction was removed, and membrane vesicles were harvested by centrifuging at 40,000 × g at 4°C for 20 min.

Tests for calcium binding protein. (i) Stains-all. The cationic carbocyanide dye Stains-all, 1-ethyl-2-[3-(1-ethylnaptho[1,2-d]thiazolin-2-ylidene)-2-methylpropenyl]naptho[1,2-d]thiazolium bromide (Eastman Organic Chemicals), was used as described by Campbell et al. (8). Crude extracts from B. subtilis were subjected to electrophoresis on discontinuous acrylamide-SDS microslab gels as described above (see "Western blot assay"). The slab gels were fixed in 25% isopropyl alcohol overnight. After fixation, the gels were washed extensively with 25% isopropyl alcohol to remove SDS. The gels were then stained in the dark for 48 h with a mixture of 0.0025% Stains-all, 25% isopropyl alcohol, 7.5% formamide, and 30 mM Tris-HCl (pH 8.8).

(ii) Ruthenium red. Ruthenium red (Sigma Chemical Co.) was used as described by Charuck et al. (10). Crude extracts were electrophoresed and electroblotted onto a BA-83 nitrocellulose membrane (Schleicher & Schuell) as described before. Nitrocellulose transfers were verified by staining with Ponceau S. The transferred proteins were then fixed with 0.2% glutaraldehyde for 15 min at room temperature. The nitrocellulose membrane was stained with 25 mg of ruthenium red/liter in a mixture of 60 mM KCl, 5 mM MgCl₂, and 10 mM Tris-HCl (pH 7.5) for 10 min.

(iii) Autoradiography with ⁴⁵Ca. Autoradiography was done according to the method of Maruyama et al. (28). Crude extracts were electrophoresed and electroblotted as described before. After glutaraldehyde fixation, the membrane was soaked in a solution containing a mixture of 60 mM KCl, 5 mM MgCl₂, and 10 mM imidazole-HCl (pH 6.8). The presence of magnesium ions and the rather low pH of the buffer are designed to prevent nonspecific binding of calcium. The buffer was changed three times to remove the electrode buffer, and the membrane was incubated in the same buffer containing 1 mCi of [⁴⁵Ca]CaCl₂/liter for 10 min. After incubation, the membrane was rinsed several times with distilled water. Excess water was absorbed by pressing the membrane between two sheets of Whatman paper no. 1. After the membrane had been dried at room temperature for 2 to 3 h, it was exposed to Kodak Xar-5 X-ray film for 2 to 3 days.

Time course of calmodulinlike protein production. Three liters of B. subtilis 3036 culture was grown in CDSM as previously described. A total of 12 flasks were used. Each flask contained 250 ml of CDSM. The flasks were harvested in duplicate at different stages of the sporulation cycle (t₀, t₁, t₂, t₄, t₆, and t₈), where the subscript denotes hours after the end of log-phase growth. The cells from each flask were centrifuged at 3,000 × g for 10 min. Each pellet was resuspended in lysis buffer (2 M urea, 60 mM 2-mercaptoethanol, 1 mM CaCl₂, 2 mM PMSF, added immediately before use) and stirred for 4 h at room temperature. The lysed samples were centrifuged as before, and the supernatant fractions were collected and dialyzed overnight against 10 mM Tris-1 mM CaCl₂ (pH 7.5) at 4°C. The pH of each supernatant fraction was adjusted to 4.3, and the fraction was stirred for 30 min in an ice bath and centrifuged at 15,000 × g. Both supernatant and pellet fractions were collected and analyzed for calmodulinlike protein by ELISA and phosphodiesterase stimulation activity.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Specificity of commercial anticalmodulin antibodies. Previous studies with commercial polyclonal antibodies raised against bovine brain calmodulin showed that these antibodies appeared to be quite specific in detecting the calmodulinlike protein (BsCaM) from B. subtilis cells (18, 19). Since a major aim of the present study was to establish the cellular location of the BsCaM using these antibodies, it was crucial to establish that they exhibited specificity in relatively crude extracts of B. subtilis cells. Consequently, using crude or partially purified extracts of B. subtilis cells, we have compared the mobility on SDS gels of the putative BsCaM detected by Western blotting using the commercial antibodies, with the mobility of calcium-binding proteins assayed by three independent detection procedures.

View larger version (79K): [in this window] [in a new window]   FIG. 1.   Western blot analysis to detect calmodulin in extracts of B. subtilis cells. Partially purified acid precipitate (protocol I of Materials and Methods) was electrophoresed on an SDS-polyacrylamide gel and blotted to nitrocellulose membrane as described in Materials and Methods. The membrane was reacted with anticalmodulin antibodies from bovine testes developed in sheep followed by anti-sheep IgG conjugated to alkaline phosphatase. Nitroblue tetrazolium-BCIP was used as a substrate to visualize the reaction. Lanes: 1, molecular mass standards phosphorylase b (112 kDa), BSA (84 kDa), ovalbumin (53 kDa), carbonic anhydrase (35 kDa), soybean trypsin inhibitor (29 kDa), and lysozyme (21 kDa); 2, bovine brain calmodulin (20 µg); 3, acid precipitate; 4, acid precipitate (5 mM EGTA); 5, bovine brain calmodulin (5 mM EGTA).

View larger version (59K): [in this window] [in a new window]   FIG. 2.   Stains-all detection of proteins following electrophoresis of crude extracts of B. subtilis. B. subtilis cells were grown in CDSM to the end of log phase, harvested, and treated as described for protocol I in Materials and Methods, up to the heating step. Alternatively, harvested cells were resuspended in 50 mM Tris-HCl (pH 7.5) and broken by passage through a French pressure cell, centrifuged, and dialyzed against a mixture of 10 mM Tris-HCl (pH 7.5), 1 mM CaCl2, and 2 mM PMSF. Samples were loaded onto 15% polyacrylamide gels containing SDS and were electrophoresed and stained with Stains-all as described in Materials and Methods. Lanes: 1, molecular mass standards myosin (203 kDa), -galactosidase (118 kDa), BSA (86 kDa), ovalbumin (51.6 kDa), carbonic anhydrase (34.1 kDa), lysosyme (19.2 kDa), and aprotinin (7.5 kDa); 2, calmodulin from bovine brain; 3 and 4, B. subtilis crude extracts lysed with urea as described in Materials and Methods; 5, extracts prepared from cells lysed by French pressure cell treatment.

Time course of calmodulin production. Earlier work (18) described a time course of production of the calmodulin during growth and sporulation of B. subtilis cells, but it was based on quantifying only phosphodiesterase-stimulating activity following four purification steps, including a column chromatography step. We reexamined the time course by using the relatively crude acid-precipitated fractions of extracts of B. subtilis in a quantitative ELISA with commercial antibodies (Fig. 3). Phosphodiesterase stimulation assays of these same acid-precipitated fractions gave a pattern nearly identical to that seen in Fig. 3 except that the stimulation caused by the t₆ sample was about the same as that seen at t₂, rather than showing the lowest activity for BsCaM (data not shown). By both assays, the maximal activity of BsCaM is seen at the end of log-phase growth and decreases by less than twofold over 8 h. We conclude that the calmodulinlike protein of B. subtilis is produced constitutively and does not change much during sporulation.

View larger version (17K): [in this window] [in a new window]   FIG. 3.   Measurement of BsCaM in sporulating cell by ELISAs. B. subtilis cells were grown in CDSM to t12, 12 h after the end of log growth. Samples collected at different stages of sporulation (t0, t2, t4, t6, and t8) were partially purified as described for protocol II in Materials and Methods. The acid-precipitated fraction was reacted with anticalmodulin antibodies. B. subtilis BsCaM was detected by ELISA as described in Materials and Methods. Color development was detected in a Bio-Tech Plate Reader at 405 nm.

Presence of calmodulin in membrane vesicles. Numerous observations have suggested that, in eukaryotic cells, calmodulin is found throughout the cytoplasm, in association with microtubules of the mitotic apparatus, and at cellular membranes depending on cell type and developmental state (52). Our observations (cited above) that more calmodulinlike activity could be extracted from cells when urea and 2-mercaptoethanol were included in the buffers were consistent with a BsCaM being associated with cell membranes. To explore this possibility, we prepared protoplasts of B. subtilis cells by using lysozyme and isolated the membrane fraction by differential sedimentation. This membrane preparation gave a strongly positive test against commercial anticalmodulin antibodies (Table 2) in a standard enzyme-linked immunoassay. When vesicle preparations were submitted to electrophoresis on 15% acrylamide-SDS gels, blotted onto nitrocellulose membranes, and exposed to anti-bovine brain, anticalmodulin antibodies, a single band with an apparent molecular mass of 39 kDa was detected (data not shown). These results suggest the protein detected here in association with the membrane is the same protein described above in Table 1.

Immunocytochemical localization of B. subtilis calmodulin. In view of the evidence that the bovine antibody against calmodulin appeared to be specific for BsCaM, we used it to try to localize the calmodulin in vegetative and sporulating cells of B. subtilis. Fresh cells from various stages of growth or sporulation were harvested by centrifugation and immediately fixed as described in Materials and Methods. Thin sections of cells were fixed and prepared by standard methods and exposed to primary anticalmodulin antibodies and to a secondary antibody labeled with colloidal gold and imaged with an electron microscope. Consistent with the experiments with vesicles, we found that the gold particles were localized at the cell envelope (Fig. 4a). Fixed spores also showed a preponderance of gold particles at the outer surfaces (Fig. 5). These results are consistent with reports that appreciable quantities of calmodulin can be associated with membranes in both animal and plant cells (12, 20).

View larger version (97K): [in this window] [in a new window]   FIG. 4.   Immunoelectron microscopic localization of BsCaM in vegetative cells. B. subtilis cells were grown in CDSM to late log phase, harvested by centrifugation, and immediately fixed in 4% paraformaldehyde-0.1% glutaraldehyde and embedded in LR white as described in Materials and Methods. Thin sections mounted on nickel grids were reacted with primary antibody, anticalmodulin antibodies from bovine testes (Fig. 4a) or normal goat serum (Fig. 4b), and then with secondary antibody (rabbit anti-goat IgG conjugated to 5-nm colloidal gold) as described in Materials and Methods. Colloidal gold is seen as electron-dense particles.

View larger version (124K): [in this window] [in a new window]   FIG. 5.   Immunoelectron microscopic localization of BsCaM in B. subtilis spores. B. subtilis cells were grown in CDSM to t20. Spores were collected by centrifugation and immediately fixed for electron microscopy as described in Materials and Methods. Thin sections were treated with anticalmodulin antibodies (Fig. 5a) or normal goat serum as the primary antibody (Fig. 5b) followed by anti-goat IgG conjugated to 5-nm colloidal gold as described in Materials and Methods. The colloidal gold is seen as electron-dense particles.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

To test the hypothesis that BsCaM might be involved in controlling calcium ion movement in B. subtilis, we undertook studies on the localization of the protein with antibodies. We have demonstrated here that the bovine brain anticalmodulin antibodies are highly specific towards the BsCaM, that they reveal the presence of BsCaM in isolated membrane preparations, and that gold-labeled antibodies are associated with the envelope of vegetative cells and with the outer surfaces of the spores, perhaps with remnants of the mother cell membrane. These observations are consistent with the reports of an acidic, heat-stable, calcium ion-binding protein of 24 kDa can be purified from spores of B. cereus (48) and with our own observations of the presence of a BsCaM in B. subtilis spores (4a). In bacteria, the isolation procedures most workers have adopted suggest a cytoplasmic location of the calmodulinlike proteins (38); this has been confirmed only by immunolocalization in the case of Anabaena variabilis (41). In fact, three species of Bordetella have been reported to excrete a 10-kDa calmodulinlike protein into the culture medium (33); as these authors note, such a location has implications for activating the adenylate cyclase associated with the pathogenesis of B. pertussis, as it would for Bacillus anthracis. However, more recently, Nagai and coworkers have reported that this "calmodulinlike" protein has strong sequence similarities to an E. coli acyl carrier protein (34).

Further evidence for the presence of calmodulinlike proteins in Bacillus species has come from observations with calmodulin antagonists. Trifluoperazine and W-7 blocked germination of B. cereus spores at a stage immediately after the loss of heat resistance (48), and promethazine specifically inhibited sporulation in B. subtilis cells at stages through t₂ (6). The effects of these pharmacological agents must be interpreted with caution, especially since selection of a strain of E. coli which was resistant to trifluoperazine resulted in the identification of a gene for a malonyl-coenzyme A transacylase rather than a calmodulinlike protein (4).

In the present studies, we have used buffers containing urea and 2-mercaptoethanol to increase the recovery of the BsCaM. With these, we have found the protein to migrate at 35 to 39 kDa in standard SDS gels rather than the very weakly staining band at 23 to 25 kDa previously described (19). As Fig. 1 shows, a 38-kDa band diminished markedly when the protein was pretreated with EDTA, and a new band appeared at 17 to 18 kDa. With some stains the band at 17 kDa was evident even without treatment with chelator (Table 1). We do not have a satisfactory explanation for these differences in the electrophoretic migrations apparently caused by altering the extraction procedures. With the current method, the electrophoretic mobilities observed were highly reproducible, and the large mobility shift caused by the chelator suggests a possible dimer to monomer conversion. Sarma and coworkers (46) have found that the circular dichroism spectrum of the calmodulinlike protein from M. phlei suggests the free protein contains 52% -sheet and only 20% -helix, which shifts to 42% -pleated sheet and 46% -helix when 10 mM calcium ion is added. In contrast, eukaryotic calmodulins contain 30 to 45% -helix and 15 to 20% -pleated sheet structure (23). Such dramatic changes in secondary structure, if they occur in the BsCaM, might be the source of the electrophoretic mobility shifts caused by EDTA and those observed when the BsCaM was reacted with rhodamine isothiocyanate (19).

In searching for a protein which might be involved in the chemorepellant response of B. subtilis cells to calcium ions, Tozzi and coworkers (50) isolated a soluble [⁴⁵Ca]CaCl₂ binding protein with an electrophoretic mobility in SDS gels of 38 kDa, which after heating and treatment with nuclease converted to a 17 to 18 kDa calcium ion-binding protein. In this connection, Matsushita and coworkers had shown (30) that several compounds known to block calcium ion channels in higher organism, including -conotoxin, selectively blocked chemotactic responses of B. subtilis cells to L-alanine without altering either growth or motility. Thus, it is possible that the BsCaM is the same protein as that isolated by Tozzi et al. (50) and could have a role in controlling calcium ion flow related to chemotaxis. Definitive assignment of various possible roles for the calmodulinlike proteins will be possible only when the genes are cloned, work which is under way in our laboratories.