Previous Article | Next Article ![]()
Journal of Bacteriology, June 2008, p. 4095-4099, Vol. 190, No. 11
0021-9193/08/$08.00+0 doi:10.1128/JB.00187-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

April Jung,
and
Beth A. Lazazzera*
Department of Microbiology, Immunology, and Molecular Genetics, David Geffen School of Medicine, University of California, Los Angeles, 405 Hilgard Avenue, Los Angeles, California 90095
Received 6 February 2008/ Accepted 16 March 2008
|
|
|---|
|
|
|---|
Knowing that five extracellular signaling peptides stimulate ComA activity raises the question of whether the different signaling peptides provide different information for B. subtilis. Genes activated by ComA include those involved in extracellular and membrane functions and regulators of development, including genetic competence and sporulation (10). At least two of these peptides, ComX and CSF, by virtue of their secretion into the growth medium, activate these processes in response to high cell density (5, 17, 20, 26). ComX also appears to signal whether there are other B. subtilis organisms of the same strain present in the environment. Genetic polymorphisms were found for different strains of B. subtilis and its close relative, Bacillus mojavensis, in genes involved in producing and responding to the ComX pheromone, comQXP (1, 2, 28, 29). Each strain produces a ComX that specifically activates ComA-controlled gene expression in that strain; however, in some cases, a limited set of pheromones from other strains partially activate or inhibit ComA-controlled gene expression (2, 28). Since ComX is a strain-specific signaling molecule, this raised the question of whether other ComA-activating signal peptides exhibited strain-specific variation. Here, we show that CSF is a conserved signaling peptide produced by B. subtilis and other closely related species and that CSF can partially compensate for the lack of ComX pheromone communication between strains.
Conservation of rapC-phrC from Bacillus strains producing different ComX pheromones. To determine the CSF pherotypes of B. subtilis and B. mojavensis strains, we sequenced the rapC-phrC operon, encoding CSF and its cytoplasmic receptor RapC, from six strains previously shown to have different comQXP sequences (28) (Table 1). Chromosomal DNA was isolated from B. subtilis strains RO-OO-2, RO-FF-1, and RS-B-1 and B. mojavensis strains RO-B-2, RO-H-1, and RO-C-2 by using a standard protocol (9). The rapC-phrC locus was then PCR amplified using Taq polymerase (Qiagen) and primers that corresponded to the sequence of the B. subtilis 168 strain. For each strain, three independent PCR products were sequenced from two independent chromosomal DNA preparations.
|
View this table: [in a new window] |
TABLE 1. Strains used in this study
|
70% identity (2, 28). None of the differences were predicted either to affect the function of rapC-phrC (Fig. 1) or to change the region corresponding to the ComA-binding box in the promoter region of rapC, which suggests that this promoter would be similarly regulated in these strains. Most of the amino acid differences in RapC and PhrC across the strains were conservative and would not be predicted to affect the function of these proteins. In particular, the sequence of the mature CSF peptide was conserved among all of the strains (Fig. 1B), and all of the PhrC protein variants were predicted by the SignalP 3.0 program to have a functional signal sequence and signal peptidase cleavage sites (7). These data suggest that for strains that produce different ComX pheromones, they all produce identical CSF peptides.
![]() View larger version (15K): [in a new window] |
FIG. 1. Locations of amino acid differences in the RapC and PhrC proteins of B. subtilis and B. mojavensis. (a) Black dots show the locations of amino acid differences. Below these dots are the amino acids found at these positions. The positions of these substitutions are shown relative to the known structural feature of RapC. (b) Linear amino acid sequence of PhrC. Where there are amino acid differences, two or three amino acids are listed. The amino acid shown above the linear sequence is that found in B. subtilis strains. The amino acids shown below the linear sequence are those found in B. mojavensis strains. Arrows indicate putative signal peptidase cleavage sites as determined by SignalP 3.0 (7). The sequence of the mature CSF peptide is boxed.
|
![]() View larger version (26K): [in a new window] |
FIG. 2. Wild-type isolates of B. subtilis secrete CSF. Culture supernatants from B. subtilis wild isolates and their phrC derivatives were tested for the ability to induce ComA-controlled gene expression in the BAL188 (JH642 srfA-lacZ) reporter strain. The increase in induction was calculated as the level of β-galactosidase specific activity for cells treated with culture supernatant divided by the level of activity for untreated cells. Plotted is the average increase in induction from three independent experiments with the standard error of the mean. Filled and open symbols correspond to results for culture supernatants from strains that either have the gene for CSF or do not ( phrC), respectively. Conditioned medium was from the following B. subtilis strains: RO-OO-2 (circles), RO-FF-1 (squares), and RS-B-1 (triangles).
|
phrC mutation was introduced into the RO-OO-2, RO-FF-1, and RS-B-1 strains. Genetically competent cells of these strains were prepared by growing the strains in a defined minimal medium (14) to an optical density at 600 nm (OD600) of 2 to 3; these cells were then transformed with chromosomal DNA from BAL941 (
phrC::tet), using standard protocols (12). Culture supernatants from these
phrC derivatives were assayed, as described above, and were found to be unable to induce srfA-lacZ expression (Fig. 2). These data indicate that phrC, and by extension CSF, was required for the inducing activity produced by the wild B. subtilis strains.
Effect of CSF on the quorum response of cultures of mixed-ComX-pherotype cells.
To determine the role of CSF in the context of a mixed community producing different ComX pheromones, we created cultures in which 10% of the cells were of a "reporter strain" that produced and responded to ComXJH642 and had a srfA-lacZ fusion. The remaining 90% of cells were of a "neighbor strain" that produced ComXJH642, ComXRS-B-1, or no ComX pheromone. To create isogenic neighbor strains, either the comQXP region from JH642 or RS-B-1 or a heterologous gene was introduced into the chromosome at the thrC locus. The comQXP locus was PCR amplified, from 200 bp upstream of comQ to 11 bp downstream of comP, and then cloned into the BamHI and EcoRI sites of the thrC integration vector pRDC19 (3). The resulting plasmids, pBL58 [thrC::(comQXPJH642 erm)], pBL220 [thrC::(comQXPRS-B-1 erm)], and pRDC19 [thrC::(xylAp-xylR erm)], were each transformed into BAL218, and transformants in which thrC was replaced with the indicated alleles were selected. To prevent ComX pheromone production from the native locus, the
comQ::spc allele was backcrossed into these strains.
The reporter and neighbor strains were grown separately in defined minimal medium with shaking at 37°C to an OD600 of
0.1, at which point the reporter strain and one of the neighbor strains were mixed. β-Galactosidase specific activity was measured during growth of the mixed culture, and the percentage of the culture that was comprised by each of the two strains was measured at the end of growth as the number of CFU per ml on the appropriate selective agar plates.
When the reporter strain was grown in the presence of a neighbor strain producing ComXJH642, the expression of srfA-lacZ was induced at an OD600 of
0.2 (Fig. 3A). In contrast, srfA-lacZ expression was induced two cell doublings later when the reporter strain was grown in the presence of a strain that either did not produce ComX (n = 3; P < 0.01) or produced ComXRS-B-1 (n = 3; P < 0.05) (Fig. 3A). These data confirm that the laboratory strain, JH642, does not respond to the ComX pheromone from the RS-B-1 strain. They further indicate that the inability to sense the presence of cells from other strains of B. subtilis via the ComX pheromone delays the expression of ComA-controlled genes when cells are grown in a mixed community.
![]() View larger version (16K): [in a new window] |
FIG. 3. CSF partially compensates for the lack of ComX pheromone communication and leads to the induction of ComA-controlled gene expression at a lower cell density. Mixed cultures in which 10% of the cells were of a reporter strain that had a srfA-lacZ fusion and 90% were of a neighbor strain producing a different complement of extracellular signaling peptides were grown. β-Galactosidase specific activity measurements were normalized by dividing the level of β-galactosidase specific activity for each point by the highest level of β-galactosidase specific activity obtained for the group of strains assayed in one experiment. The graphs shown are representative of three independent experiments. (A) For all three cultures, all cells produced CSF, and BAL188 (srfA-lacZ ComXJH642) was the reporter strain. The neighbor strain either produced ComXJH642 (BAL1391) (open circles), ComXRS-B-1 (BAL139) (filled circles), or no ComX pheromone (ComX–; BAL1397) (filled squares). (B) BAL1416 (srfA-lacZ phrC ComXJH642) does not produce CSF and was the reporter strain. The neighbor strain produced ComXRS-B-1 and was either phenotypically CSF+ (BAL1392) (filled circles) or CSF– (BAL1419) (open squares).
|
phrC mutation into the reporter strain (BAL188) and the neighbor strain expressing ComXRS-B-1 (BAL1419) (Table 1). Two mixed cultures were then set up, in which the reporter strain, which cannot produce CSF (BAL1416), was grown with either a neighbor strain that expressed comQXPRS-B-1 and CSF (BAL1392) or a neighbor strain that expressed comQXPRS-B-1 but not CSF (BAL1419) (Fig. 3B). For the culture in which CSF was not produced, srfA-lacZ expression was induced at 0.5 cell doublings later than it was induced in the culture that produced CSF (n = 3; P < 0.02). These data indicate that CSF mediates communication between different ComX pherotype strains and leads to the induction of ComA-controlled gene expression at a lower cell density than can be achieved without CSF. CSF only partially compensated for the lack of ComX pheromone communication, as ComA-controlled gene expression was delayed when strains of different ComX pherotypes were grown together, even in the presence of CSF. Conclusions and implications. The reason that B. subtilis produces multiple signaling molecules that converge to regulate the quorum response is unclear. In light of the strain specificity exhibited by ComX pheromone, we set out to determine whether the CSF signaling molecule was also strain specific. Our data indicate that CSF is not; it is a conserved signaling molecule produced by strains of B. subtilis and the closely related species B. mojavensis. Furthermore, CSF can mediate communication between strains that are unable to communicate via ComX. This is the first evidence that a bacterium that exhibits strain level specificity for a secreted signaling molecule produces an additional signaling molecule to allow communication between strains. Strain-specific variation has also been observed for the extracellular signaling peptides that induce genetic competence in Streptococcus pneumoniae (31) and virulence gene expression in Staphylococcus aureus (15) and Bacillus cereus (24), and it will be interesting to learn whether these bacteria also produce a species-specific signaling molecule, as does B. subtilis.
It is unknown why bacteria secrete multiple signaling molecules that differ in the ability to communicate the relatedness of the surrounding bacteria. In addition to the fact that B. subtilis produces strain- and species-specific signals, Vibrio harveyi produces species- and genus-specific signals, as well as a universal signal (6, 13, 32). The fact that bacteria as distantly related as V. harveyi and B. subtilis exhibit this phenomenon suggests that this may be a common property of signaling molecules produced by many bacterial species.
To understand why B. subtilis produces both strain- and species-specific signaling molecules, it will be necessary to discern the selective pressure that has led to the strain level of variation in ComX pheromone. Since ComX pheromone is required for the development of genetic competence (i.e., the natural ability to take up DNA), one consequence of ComX pheromone variation is sexual isolation from different strains (25). Therefore, it has been proposed that by B. subtilis cells responding only to ComX pheromone from genetically identical clones, these cells avoid the uptake of DNA from closely related, but not genetically identical, strains that could carry harmful genetic elements (1). If sexual isolation is the fitness-enhancing factor that has selected for the diversification of the ComX pheromone, then what advantage does communication via CSF provide for strains of different ComX pherotypes? Perhaps B. subtilis has to strike a balance between sexual isolation and cooperation for other processes activated by ComX and CSF through the ComA transcription factor (e.g., surfactin or secreted pectate lyase production) (10). And so, although B. subtilis possibly excludes harmful genetic variation via the lack of ComX communication, it can still gain the benefits of cooperation by signaling via the conserved molecule CSF. If CSF accumulates to sufficient levels to be sensed by B. subtilis cells, which is dependent on the accumulation of ComX itself (17), the likelihood of the existence of a high density of cooperative cells may be sufficiently great to make induction of the quorum response possibly beneficial. Clearly, there are many experiments yet to be done to understand the potential benefits offered to a bacterium that can communicate at both the strain and species levels.
Nucleotide sequence accession numbers. The GenBank accession numbers for the obtained sequences are as follows: AY664811 (B. subtilis RO-OO-2), AY667584 (B. subtilis RO-FF-1), AY667585 (B. subtilis RS-B-1), AY667586 (B. mojavensis RO-B-2), AY672642 (B. mojavensis RO-H-1), and AY672643 (B. mojavensis RO-C-2).
This work was supported by NIH Public Health Service grant AI48616 to B.A.L. M.P. was supported in part by USPHS National Research Service Award GM07185.
Published ahead of print on 28 March 2008. ![]()
Present address: XLTech Group, 1901 S. Harbor City Blvd., Ste. 600, Melbourne, FL 32901. ![]()
Present address: St. Mary Medical Center, 1050 Linden Ave., Long Beach, CA 90813. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»