OUP user menu

Peptide pheromone induced cell death of Streptococcus mutans

Fengxia Qi , J. Kreth , Celine M. Lévesque , Olga Kay , Richard W. Mair , Wenyuan Shi , Dennis G. Cvitkovitch , Steven D. Goodman
DOI: http://dx.doi.org/10.1016/j.femsle.2005.08.018 321-326 First published online: 1 October 2005

Abstract

Streptococcus mutans is considered one of the main causative agents of human dental caries. Cell–cell communication through two-component signal transduction systems (TCSTS) plays an important role in the pathogenesis of S. mutans. One of the S. mutans TCSTS, ComDE, controls both competence development and biofilm formation. In this study, we showed that addition of exogenous competence-stimulating peptide (CSP) beyond the levels necessary for competence inhibited the growth of S. mutans in a ComDE-dependent manner. We also demonstrated that further increases of CSP stopped S. mutans cell division leading to cell death. Use of CSP as a possible therapeutic agent is discussed.

Keywords
  • Streptococcus mutans
  • Competence stimulating peptide
  • Programmed cell death
  • Mutans
  • Biofilms

1 Introduction

Streptococcus mutans is considered the major etiologic agent involved in human dental caries [1,2]. S. mutans utilizes quorum sensing signaling systems to regulate the expression of several virulence factors, including competence development, acid resistance, and biofilm formation [35]. One of the S. mutans quorum sensing systems involves at least three components, a competence stimulating peptide (CSP) and a two-component signal transduction system (TCSTS) called ComDE. CSP, encoded by comC, is likely synthesized as a precursor, which is then believed to be processed and exported by the ABC transporter CslAB (ComAB in S. mutans) [6]. Based on the model proposed for the homologous system in Streptococcus pneumoniae when the mature CSP peptide reaches a threshold concentration, it binds to the membrane-bound histidine kinase sensor ComD, causing its autophosphorylation. Phosphorylated ComD then transfers the phosphate group to its cognate intracellular response regulator ComE, which then activates transcription of comX encoding an alternative sigma factor. ComX enables transcription of several late competence-related genes involved in DNA uptake and integration [7,8]. Since CSP also seems to induce competence-related stress responses [9,10], we hypothesized that levels of CSP that exceed that needed for the development of competence (an overdose) could invoke a preprogrammed response shifting the cell's metabolic balance and causing a detrimental effect. In this study, we tested this hypothesis in both planktonic and biofilm cultures and demonstrated that an overdose of CSP inhibited cell division leading to cell death in S. mutans. The potential to use CSP as a stand alone or adjunct therapy is discussed.

2 Materials and methods

2.1 Bacterial strains and culture

All S. mutans strains were grown anaerobically (80% N2, 10% CO2, and 10% H2) at 37 °C in Todd–Hewitt (TH) broth (Difco) or on TH agar plates. When needed, erythromycin was added at 10 μg per ml. S. mutans wild-type strains UA140 [11] and UA159 [12] were used for this study. The S. mutans comD, comE, and comX null mutants were first constructed in UA159 by a PCR-based deletion strategy involving restriction–ligation and allelic replacement to avoid polar effects on the downstream genes [13]. The primers used to construct and confirm the mutants are listed in Table 1. The S. mutans UA140 comD, comE, and comX null mutants were then obtained by transferring the individual mutations to UA140 using the PCR products generated with the primer pairs comD-P1/comD-P4, comE-P1/comE-P4, and comX-P1/comX-P4, respectively. The transformants were confirmed by PCR. Construction of strain UA140 comC was described previously [14].

View this table:
1

Primers used to construct the comD, comE, and comX null mutants by PCR restriction-ligation mutagenesisa

PrimerNucleotide sequenceb (5′–3′)
comD-P1CACAACAACTTATTGACGCTATCCC
comD-P2GGCGCGCCAACTGGCAACAGGCAGCAGACC
comD-P3GGCCGGCCTCAAAACGATGCTGTCAAGGG
comD-P4AGATTATCATTGGCGGAAGCG
comE-P1GCTGGTTATCGTTTGGTGC
comE-P2GGCGCGCCGCAATGGTGGTTTCAAGACG
comE-P3GGCCGGCCACTTTTTCAGTGCCATCGC
comE-P4ATCAAGCAACTCCATCTCAGG
comX-P1AACACAGCAGTTAAGCCCTAGC
comX-P4GGTTCTACAATTTCACCTTTACCTG
  • aPrimers were designated and analyzed with MacVector 7.2 software.

  • b Asc I restriction sites are in boldface; Fse I restriction sites are underlined.

2.2 CSP assays

To determine the minimum inhibitory concentration (MIC) of CSP to inhibit S. mutans growth, log phase cells were diluted to approximately 105 cfu/ml (colony forming unit per milliliter) in TH broth and added to 96-well plates (100 μl/well). CSP was 2-fold serially diluted, added to the cell suspensions, and the plates were incubated overnight at 37 °C under anaerobic conditions. The MIC was determined by the lowest concentration of CSP required to inhibit bacterial growth.

To determine the effect of brief CSP treatment on S. mutans cell growth, a pulse assay was performed. Overnight cells were diluted 1:10 into fresh TH broth and incubated anaerobically at 37 °C for 2 h. The cells were then collected by centrifugation and resuspended into fresh TH broth to a final concentration of approximately 109 cfu/ml and containing different concentrations (0.4–4 mM) of CSP. After an incubation of 10 min at room temperature, the cells were collected by centrifugation, resuspended into fresh TH broth to a final concentration of approximately 107 cfu/ml, and incubated at 37 °C. Cell growth was monitored for 4 h at 2-h intervals by optical density (OD) readings at 600 nm.

2.3 Biofilm growth and live/dead cell assay

For biofilm formation, cells were grown anaerobically in TH overnight, and diluted 1:100 (final cell density of ∼106 cfu/ml) into TH containing 0.5% sucrose in the Lab-Tek® II Chamber Slide™ System (Nalge Nunc International; Naperville, IL). After initial attachment for 3 h at 37 °C, CSP was added to a final concentrations 5, 50, or 100 μM. The cells were further incubated anaerobically overnight at 37 °C. Planktonic cells and biofilms were stained by replacing the medium with LIVE/DEAD Bac Light bacterial cell stains diluted in PBS according to manufactures instructions (Molecular Probes; Eugene, OR). Confocal Laser Scanning Microscopy was performed using LSM 5 PASCAL with LSM 5 PASCAL software (Carl Zeiss) equipped with detectors and filter sets for monitoring red and green fluorescence. Images were obtained with a 40 × 1.4 Plan-Neofluar and a 63 × 1.4 Plan-Neofluar oil objective.

3 Results

3.1 Inhibition of cell growth by CSP

To assay the effect of CSP on S. mutans cell growth, an MIC test was first performed using 96-well plates. MIC values varying between 0.65 and 1.3 μM and 10 and 20 μM were obtained for wild-type strains UA140 and UA159, respectively (Table 2). To test the specificity of CSP on growth inhibition, closely related streptococcal species such as Streptococcus sobrinus OMZ176, Streptococcus sanguinis NY101, and Streptococcus gordonii ATCC10558 were also included in the MIC test. As expected, these streptococcal species were not inhibited by S. mutans CSP even at the highest concentration used (65 μM) (Table 2).

View this table:
2

MIC of CSP against S. mutans and other streptococcal species

Species/strainsMIC (μM)
S. mutans UA1400.65–1.3
S. mutans UA15910–20
S. sobrinus OMZ176>65
S. sanguinis NY101>65
S. gordonii ATCC10558>65
UA140 comC null0.65–1.3
UA140 comD null>65
UA140 comE null>65
UA140 comX null0.65–1.3

The profound and specific effect that CSP had on growth inhibition suggested that it might be a useful treatment for selective elimination of S. mutans from the oral cavity. Although we have not as of yet determined the half life of CSP under our experimental conditions nor the oral cavity, we were concerned that exogenously added CSP could turnover before it had any growth effect. In order to determine whether CSP exposure could inhibit S. mutans cell growth over a short period, a pulse assay was developed. Briefly, CSP is added to S. mutans for a brief period and subsequently washed away. Each assay was performed multiples times with the two different wild-type strains (see Section 2). A typical result is shown in Fig. 1. At 4-h post-treatment the cell growth of UA140 (Fig. 1A) and UA159 (Fig. 1B) was inhibited by approximately 50% when the cells were pulsed with 0.4 mM CSP, and completely inhibited with 1.6 mM CSP. When the incubation was prolonged, the cell growth started to recover, and at 16-h post-treatment no difference could be observed between the CSP-treated and untreated cells (data not shown). To see if this pulse inhibitory effect was species specific, other streptococcal species such as S. sobrinus and S. sanguinis were also tested in the assay and found not to be affected by CSP addition (in fact, their growth appeared to be enhanced probably due to the extra nutrients provided by the peptide. Data not shown). These results suggested that a short period of CSP treatment caused a specific growth arrest in S. mutans.

1

Temporary growth inhibition of S. mutans wild-type strains UA140 (A) and UA159 (B) by CSP. The experiment was performed in a 96-well plate with triplicate wells for each treatment. Cell density was measured by using a plate reader (BioRad). The experiment was repeated at three times. Numbers represented the average of the triplicate samples from one representative experiment. Variations between experiments were within 20%.

3.2 The inhibitory effect of CSP is mediated through the ComDE pathway

The involvement of the TCSTS ComDE in the CSP inhibition of cell growth was determined by assaying the S. mutans UA140 comC, comD, comE, and comX null mutants. An MIC test on these mutant strains demonstrated that cell growth of the comD and comE null mutants were no longer inhibited by CSP (Table 2). As expected, the comC null mutant exhibited the same response as the wild-type strain UA140. Surprisingly, the comX null mutant also displayed a response indistinguishable from the wild-type (Table 2). Similar results were also obtained with the pulse assay (data not shown). These results suggested that the cell's CSP response was indeed mediated through the ComDE pathway; however, this response was independent of the competence pathway that involved comX regulated genes.

3.3 CSP induced cell death in planktonic and biofilm cultures

To further investigate the mechanism of CSP's inhibitory effect on S. mutans growth, we examined microscopically the CSP-treated and untreated cells. Surprisingly, some (∼10%) of the CSP-treated wild-type cells appeared swelled either as individuals or in cell chains (Fig. 2A). More interestingly, in some short cell chains of 4–5 cells only 1 or 2 were swelled, which made the cell chain look like a balloon (Fig. 1A inset). This morphology suggested that cell division was inhibited in these cells. To further test whether these balloon-like structures were still viable, we stained the cells with BacLight, a live/dead stain kit that stains live cells green and dead cells red. As shown in Fig. 2B, all balloon-like cells were stained red, while normal-looking cells were stained green. This result suggested that CSP may have caused inhibition of cell division in a certain population of the cells, leading to cell death. To see if this effect was also mediated via the ComDE pathway, CSP-treated UA140 comD null mutant cells were also examined. As shown in Fig. 2C, the mutant cells did not show the morphological changes observed with the CSP-treated wild-type UA140 cells. Consequently, all of the comD null mutant cells were stained green with BacLight (Fig. 2D). As a final test 107 cfu of UA140 and UA159 were treated with CSP as described for Fig. 1, in the pulse assay and plated on TH agar. At a CSP concentration of 0.4 mM only 1% of each culture survived while at a CSP concentration of 1.6 mM no cfus were detected (data not shown). This further indicates that CSP was killing S. mutans and not simply causing growth arrest.

2

Effect of CSP on cell division and viability. (A) Phase-contrast image of CSP-treated UA140 cells. The inset is a close caption of the “balloon-like” morphology of the CSP-treated cells. (B) Fluorescent image of CSP-treated UA140 cells after live/dead staining. Arrowhead indicates the swelled cells. (C) and (D) Phase-contrast and fluorescent images of CSP-treated UA140 comD null mutant cells, respectively. (E) Confocal image of untreated UA140 biofilm cells. (F) Confocal image of UA140 biofilm cells treated with 5 μM CSP. Green: live cells; red: dead or membrane compromised cells.

To test whether CSP also induced cell death in S. mutans biofilms, UA140 biofilms were grown in the presence and absence of CSP and the cells were stained using the BacLight kit. As shown in Fig. 2E, only a small percentage (<1%) of cells were dead in the absence of CSP. In contrast, in the presence of 5 μM CSP over 50% of the cells in the biofilm were dead (Fig. 2F). Increasing the CSP concentrations up to 100 μM did not further increase the proportion of dead cells (data not shown). These results indicate that similar to competence development in S. mutans biofilms [15], only a subpopulation of cells responded to CSP regardless of the exogenous concentration.

4 Discussion

The oral mucosal surfaces are colonized by a large number of bacteria [1618]. At a homeostatic state this so called “indigenous microflora” plays important roles in human health by preventing colonization by pathogens. However, when the ecological balance is disrupted by environmental perturbations, diseases such as dental caries and periodontal infections may result from overgrowth of indigenous pathogens [2,19,20]. A unique challenge in treating dental caries and periodontal diseases is to selectively inhibit the pathogenic bacteria while leaving the protective indigenous flora unharmed. Conventional treatments of mechanical removal of dental plaque or antibiotic killing of all plaque bacteria provide only a temporary solution to the problems due to the non-sterile environment of the oral cavity. Therefore, there is a need to find antimicrobial compounds with very narrow and specific activity. In this study, we explored the possibility of using a bacterial pheromone as a potent (species)-specific inhibitor.

We analyzed the effect of CSP, a peptide signaling pheromone, on S. mutans cell growth. We showed that CSP at concentrations higher than used for competence development (∼0.5 μM) [21], inhibited both S. mutans strains UA140 and UA159 growth in a regular MIC assay, and that treatment of cells with CSP for a short period of time could lead to growth arrest. Further analyses demonstrated that CSP exerts this effect by inhibiting cell division, which ultimately leads to cell death. Mutational analyses suggested that the ComDE quorum sensing pathway mediated CSP-induced cell death in a ComX-independent manner. Taken together, these results suggested that CSP may have the potential to be developed into a species specific antimicrobial agent against S. mutans for manipulating the dental plaque ecology as a way for treating dental caries. A preliminary examination of the inhibitory effect of CSP on other S. mutans strains suggested that the effect was highly variable among strains, further indicating that there may be substantial heterogeneity among CSP-receptor interactions between strains. Additional studies are required to test the spectrum of CSP's activity against other S. mutans clinical isolates.

It is worth noting that using bacterial pheromones for species-specific inhibition of cell growth is not limited to S. mutans. Ji et al. [22] showed that a pheromone produced by one staphylococcal strain interfered with virulence factor expression of another strain, and this pheromone was proposed as a therapeutic or prophylactic initiative. Hidalgo-Grass et al. [23] demonstrated that administration of SilCR, a 17 amino acid pheromone peptide produced by the group A streptococci (GAS), to mice challenged with those bacteria resulted in clearing of the bacteria and survival of the animals. Steinmoen et al. [24] also found that addition of CSP to S. pneumoniae for competence development caused cell lysis in a subfraction of the cell population. In contrast to our observations of S. mutans cell death in lag and log phases, S. pneumoniae cell death was only observed in stationary phase; this may indicate an important distinction between the two bacterial cell death mechanisms During preparation of this manuscript, Oggioni et al. [25] reported that the pneumococcal CSP inhibited the growth of S. pneumoniae in vitro and reduced the blood count of the bacterial cells in vivo. These findings suggest that peptide pheromones can have strong effects on bacterial physiology. It has yet to be determined if the concentrations used in the assays, by ourselves and others, constitute an artificial overdosing of the normal physiologic pathways or whether these concentrations are part of a programmed pathway that may have relevance in the context of a biofilm where the local concentration of peptide is likely to be high.

Although we have made progress in demonstrating that this CSP induced phenomenon is com dependent we were unable to identify downstream genes that are the target of this signal transduction; a comX deficiency failed to yield a phenotype. While we currently pursue a microarray analysis of CSP induced/repressed genes that will lead to downstream gene candidates, we have preliminary evidence that one essential cell division gene is strongly repressed. FtsA is the structural gene of a protein required in the formation of the septum of dividing cells and is highly conserved throughout eubacteria. We have found that pulsing S. mutans with 10 μM CSP reduces expression of a ftsA transcriptional fusion by 50-fold (data not shown). Although we do not know if this is the primary mechanism through which CSP exerts its effects, transcriptional regulation of ftsA by a TCSTS has been observed in Bacillus subtilis [26]. Thus ftsA remains a good candidate for com dependent regulation.

Acknowledgements

We thank Randy Eckert for chemical synthesis and purification of CSP and Jennifer Treglown for comments on the manuscript. This work was supported in part by NIH Grants U01-DE15018, R01-DE014757, NIDCR T32 Training grant DE007296, and a Delta Dental Grant WDS78956. D.G.C. is supported by a Canada Research Chair grant, an NIDCR Grant R01 DE013230–0 and Grant MT-15431 from the Canadian Institutes of Health Research. S.D.G. is supported by NIH grant R01 DE013965.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
View Abstract