The motPS genes are downstream of the ccpA gene, which encodes a central regulator of carbon metabolism, forming a putative ccpA-motPS operon (6, 7, 16). Coordinated expression of ccpA and motPS could represent a multipronged response to alkali stress, since both Na⁺-coupled MotPS-dependent motility and increased metabolic production of acids are adaptive to high pH (11, 13). Sequence analysis of the up-motile mutant (AB::Tn-M) selected in a motAB mutant strain (AB::Tn) showed no mutations in the ccpA and motPS coding sequences (strains are listed in Table S1 in the supplemental material). However, a point mutation (G→A) was found at the 33rd nucleotide following the stop codon of the ccpA gene, within a stem-loop structure in the intergenic region between ccpA and motP that has the potential to serve as an intrinsic transcriptional terminator. The mutation is predicted by the Mfold program (22) to change the free energy (ΔG) of the RNA secondary structure from −18.5 kcal/mol to −12.3 kcal/mol. This mutation was confirmed to be sufficient to confer the up-motile phenotype after deletion of the native ccpA-motPS operon from B. subtilis AB::Tn using the method described previously by Horton (8), producing strain AB::TnΔCPS. Upon introduction of the mutant or wild-type ccpA-motPS operon into the amyE locus of this strain, the mutant ccpA-motPS locus supported the same up-motile phenotype as the original up-motile strain on soft agar plates, whereas the strain expressing the wild-type locus did not (shown in Fig. S1 in the supplemental material together with a diagram of the mutation site; primers and details of strain construction are available on request).

The levels of MotP and MotS in the membranes of the up-motile AB::Tn-M strain that lacks motAB, its AB::Tn parent strain, and the wild-type strain were analyzed by Western blots of sodium dodecyl sulfate-10% polyacrylamide gels (17) carried out using a chemiluminescence protocol according to the manufacturer's instructions (Amersham Biosciences). The polyclonal anti-MotP or anti-MotS antibodies used for detection were raised in rabbits against synthetic peptides corresponding to residues 88 to 100 of MotP (SLSDHARKHGLL) and to residues 1 to 14 of MotS (MKLRRERFERRNGS), with an additional cysteine added to the C terminus to facilitate conjugation to keyhole limpet hemocyanin (Operon Biotechnologies, Inc., Tokyo, Japan); a purified immunoglobulin G fraction (Melon Gel IgG Spin purification kit; Pierce Biotechnology, Inc., IL) was used. The intensity of the MotP and MotS bands was comparable in the wild-type and AB::Tn samples, as expected, while the levels of MotP and MotS were increased 3.8 and 2.1 times, respectively, in the AB::Tn-M samples (Fig. ​(Fig.1A1A).

Transcript levels for ccpA and motPS were measured for the motAB⁺ strains CPS and CPS-M that express the wild-type or mutant ccpA-motPS locus, respectively, only from the amyE locus, with a ΔCPS strain as a negative control. RNA was prepared as described previously (9), and Northern analysis was carried out using digoxigenin RNA probes (DIG RNA labeling kit, SP6/T7; Roche Applied Sciences). Both the motPS and ccpA probes (Fig. 1B and C) hybridized to a 2.7-kb band in RNA from both CPS and CPS-M (Fig. ​(Fig.1C).1C). This size corresponds to the expected size for ccpA-motPS mRNA and was the only band observed with the motPS probe. The amount of the 2.7-kb mRNA in CPS-M cells was about twofold higher than that in CPS cells. In addition to the 2.7-kb ccpA band, a significant amount of ccpA-containing mRNA in the CPS strain was found in two bands around 1.1 kb in size, the expected size for a monocistronic ccpA transcript. A much weaker 1.1-kb mRNA signal was observed in the CPS-M mRNA (Fig. ​(Fig.1C,1C, left). These results indicated that ccpA and motPS form an operon, since both ccpA and motPS probes hybridized to the 2.7-kb transcript. Consistent with a transcriptional termination function for the intergenic stem-loop, transcription of ccpA alone occurred at a higher level than transcription of the entire operon in the wild-type strain, whereas the level of the polycistronic ccpA-motPS mRNA is increased in the up-motile mutant, and little monocistronic ccpA mRNA was detected.

To better define the influence of the stem-loop element on transcription, ccpA-lacZ and motPS-lacZ fusions were generated in the wild-type (motAB⁺) strain using the pMutin4 integration plasmid (18) to fuse lacZ to the ccpA gene upstream of the stem-loop or to motS downstream of either a wild-type stem-loop (motPS-lacZ) or an up-motile mutant stem-loop (motPS-lacZ-M) (Fig. ​(Fig.2A).2A). The resulting strains grew comparably at 37°C in 2× TY medium (10) (Fig. ​(Fig.2B).2B). Samples were taken at different points during growth for measurements of β-galactosidase activity (4). The most striking feature of the expression patterns was that motPS-lacZ transcriptional activity was significantly lower than that of ccpA-lacZ, whereas the transcriptional activity of motPS-lacZ-M was close to that of ccpA-lacZ, with a 3.6-fold increase relative to the motPS-lacZ fusion (Fig. ​(Fig.2C).2C). These results are consistent with the observation that the up-motile mutation results in increased levels of ccpA-motPS polycistronic mRNA.

A microscopic examination was carried out on negatively stained preparations (21) of four motile strains (wild type, AB::Tn-M, ΔPS, and CPS-M) and three nonmotile strains (AB::Tn, AB::Tn-ΔPS, and ΔAB). Strains expressing motPS from the up-motile mutant ccpA-motPS locus had an average of 12 flagella/cell whether or not MotAB was also present, a number of flagella/cell that was similar to that of the wild-type strain (average of 10) and higher than the number of flagella/cell in cells expressing only MotAB (average of 7) or MotPS (average of 5) from the wild-type ccpA-motPS locus (Table ​(Table1).1). The inability of the MotPS-only cells to swim in liquid, in contrast to the MotAB-only cells, is probably due to lower numbers of Mot complexes in the former cells, and that number is increased by the up-mutation. The motAB motPS double mutant had an average of 2 flagella/cell and was the only strain in which flagellar length was also significantly shorter than that of the wild type (Table ​(Table1).1). Calvio et al. (3) recently identified the B. subtilis swrA gene of the dicistronic swrAB operon as the locus of the ifm mutation that increases flagellar number and results in hypermotility (5). We verified that up-motile strains contained no changes in the swrAB sequence. Our results support other evidence that the presence of Mot complexes influences flagellar assembly (10, 19) and indicate that the presence of either the MotAB or MotPS stator is sufficient to allow normal flagellar biogenesis in B. subtilis.

Finally, the contribution of MotPS to the swimming speed under different conditions was assessed (10) in three strains that have wild-type motAB loci but differ in motPS status: wild-type B. subtilis; CPS-M, expressing a mutant ccpA-motPS operon in the amyE locus; and ΔPS, lacking motPS. First, cells grown at pH 7.0 in TY medium (which contains 14 to 17 mM Na⁺), with or without the addition of 200 mM NaCl, were transferred into TY medium without added NaCl, at different pH values. The effect of the Na⁺ channel blocker 5-(N-ethyl-N-isopropyl)-amiloride (EIPA), which selectively inhibits MotPS-dependent swimming (11), was assayed. Indeed, EIPA did not significantly inhibit the motility of a strain lacking motPS (ΔPS) under any condition of pH or Na⁺ content, whereas inhibition was observed in the motPS-containing strains (see Fig. S2 in the supplemental material). The inhibition by EIPA as a function of pH and pregrowth with 200 mM added Na⁺ showed that MotPS has a significant role in motility in the up-motile MotPS strain even at pH 6.0 without pregrowth with added Na⁺, whereas MotPS expressed from the wild-type locus contributed to swimming at pH 6.0 only if cells were pregrown with added Na⁺ (Fig. ​(Fig.3A).3A). At all pH values, the role of MotPS, as assessed by percent EIPA inhibition, was greater in the strain expressing the up-motile ccpA-motPS locus. Next, the contribution of MotPS to swimming of the motAB⁺ strains at a low protonmotive force (lowered by protonophore carbonyl cyanide m-chlorophenylhydrazone [CCCP]) or elevated viscosity (achieved by the addition of polyvinylpyrrolidone [PVP]) was studied in cells pregrown and assayed in the presence of 200 mM NaCl at pH 8.5, conditions that maximized the MotPS contribution (see Fig. S2 in the supplemental material). Under these conditions, MotPS clearly contributed to swimming in the presence of added CCCP or PVP, with the mutant ccpA-motPS locus conferring greater adaptability than the wild-type locus to either low protonmotive force or elevated viscosity (Fig. ​(Fig.3B3B).

The swimming-speed assays show that MotPS plays a role in the motility profile of wild-type B. subtilis with a functional MotAB stator, especially once the organism is exposed to elevated Na⁺ levels. The increased impact of the up-motile MotPS phenotype in strains containing the stem-loop mutation is most evident at elevated Na⁺, pH, and viscosity or at low protonmotive force, conditions that could select for mutations of this type when “undomesticated strains” are exposed to them in the environment.