OUP user menu

Allelic variation in a peptide-inducible two-component system of Streptococcus pneumoniae

Peter Reichmann, Regine Hakenbeck
DOI: http://dx.doi.org/10.1111/j.1574-6968.2000.tb09291.x 231-236 First published online: 1 September 2000


The peptide SpiP of Streptococcus pneumoniae regulates the induction of a complex signal transduction system spiR1spiR2spiH. Distinct alleles of spiP and the receptor histidine protein kinase gene spiH were recognized in different pneumococcal clones. The spi system in strain KNR7/87 is adjacent to a bacteriocin gene cluster encoding putative double glycine-type bacteriocins, immunity proteins, and translocator proteins. A direct repeat element upstream of the spiR1 promoter and another three potential transcription start sites within the bacteriocin cluster indicate that SpiP functions as an inducing peptide for bacteriocin synthesis in S. pneumoniae.

  • Two-component system
  • Peptide pheromone
  • Bacteriocin
  • Quorum sensing
  • Streptococcus pneumoniae

1 Introduction

Secreted peptides play an important role in quorum sensing in Gram-positive bacteria. They represent signaling molecules of multicomponent regulatory systems that orchestrate the expression of genes involved in e.g. pathogenicity in Staphylococcus aureus, bacteriocin production in a variety of Lactobacillus spp., or genetic competence and transformation in Bacillus subtilis and Streptococcus pneumoniae (for review, see [1]). A common theme in these processes appears to be a post-translational processing of the signaling peptide [2]. In the case of the S. pneumoniae competence signaling peptide CSP and class II antimicrobial peptides, an N-terminal Gly-Gly cleavage motif serves as recognition sequence for dedicated ABC exporters and accessory proteins which function as maturation protease [3]. The secreted peptide acts as a signal molecule for a histidine protein kinase of a two-component system whose sensor domain typically consists of several transmembrane spanning segments. Activation of the His-Asp phosphorelay system appears to be part of the frequently observed autoregulatory synthesis of the peptide, and is essential for the cell density-dependent manifestation of distinct physiological states.

We have now analyzed the function of a S. pneumoniae peptide SpiP with the typical double glycine leader peptide of secreted biologically active peptides. It is located adjacent to but transcribed divergent from a two-component system related to the ComDE His-Asp phosphorelay [4,5] named SpiR1SpiR2SpiH in the present study. Several SpiP variants exist in different pneumococcal clones that are associated with distinct histidine kinase genes SpiH. The in vitro synthesized peptide induces expression of the spiR1spiR2spiH operon. We propose that an imperfect direct repeat located upstream of the spiR1 promoter and potential transcription units in an adjacent bacteriocin cluster serve as recognition sequence for SpiR-controlled gene expression.

2 Materials and methods

2.1 Bacterial strains and culture conditions

The following S. pneumoniae strains are members of genetically distinct clones (serotypes): R6 (non-encapsulated) [6]; 669 (19) [7]; SA9 [original No. 10760] (6A), Sp670 (6B), Sp628 (9V), 637 (23F), Fi2301r [original No. 43351] (23F) [8]; Hu11 (19A), F4 (23F), F1 (23F) [9]. Streptococci were grown in a semisynthetic medium containing 0.16% glucose supplemented with 0.1% yeast extract at 37°C without aeration [7]. Escherichia coli strains were grown in LB broth at 37°C with aeration.

2.2 Construction of lacZ reporter fusions

The upstream region of the spiR1/spiR2/spiH genes was amplified by PCR using oligonucleotides 5′-ACTTGGTCAAAGTGGAAGCGGTC and 5′-TAGACTGTTTTCTGCGTCTG covering the −558 to +38 region from the 3′-end of the BOX element to codon 13 of spiR1. The fragment was cloned first into the E. coli plasmid pCRII™ (Invitrogen, Groningen, The Netherlands), excised with BamHI and EcoRV and subcloned into BglII- and SmaI-digested pEVP3 [10] to create a lacZ transcriptional fusion. The resulting plasmid pPRO2 was transformed into E. coli InvαF′ (Invitrogen), and recombinant E. coli clones were selected at 30 μg ml−1 chloramphenicol.

2.3 Insertion duplication mutagenesis

Transformation of streptococci followed published procedures [7]. The plasmid pEVP3 and accordingly its derivative pPRO2 cannot replicate in S. pneumoniae, but contains a selectable chloramphenicol resistance marker. Chloramphenicol-resistant transformants of S. pneumoniae can only arise by homologous recombination between the insert and the chromosome, an event by which the cloned region becomes duplicated after integration of the plasmid into the pneumococcal chromosome [10]. After transformation with pPRO2, chloramphenicol-resistant colonies of S. pneumoniae R6bgaA::ermB defective in the β-galactosidase gene bgaA[11] were selected at a concentration of 3 μg ml−1 chloramphenicol. Ten independent transformants were isolated and insertion of the plasmid into the target region verified by PCR analysis.

2.4 β-Galactosidase assay

β-Galactosidase was assayed essentially according to published procedures [10]. Hydrolysis of ONPG was monitored at 30°C, and enzyme activity calculated according to Miller, using Nephelo units instead of OD620 of the culture [12]. Chemically synthesized SpiP peptides as deduced from the DNA sequence (Schneider-Mergener, Berlin, Germany) were added to growing cultures at the concentrations stated in Section 3.

2.5 DNA techniques

DNA was amplified by PCR with Goldstar Polymerase (Eurogentec, Seraing, Belgium). Sequencing was performed with the ABI Prism™ Dye Terminator Ready Reaction Kit (Perkin-Elmer). For amplification of the 4642-bp fragment covering the region from spiR to spiB, oligonucleotides 5′-CCGTCTCTTGCAGTGCCACCG and 5′- TCCAACAAGTCATGAGCTTCTCCAG were used. Restriction enzymes were obtained from NEB (Schwalbach, Germany). All DNA techniques followed standard protocols [13].

2.6 Database searching

Sequence database searches were carried out using BLAST [14] and FASTA [15]. Sequence alignments were constructed using the GCG 9 package or Clustal X 1.8. Sequences for microbial genomes were obtained from The Institute of Genomic Research website at http://www.tigr.org.

2.7 Nucleotide accession number

The spiP and spiR1/R2/H sequences of the different strains have been deposited in the EMBL Nucleotide Sequence Database under the following accession numbers: S. pneumoniae KNR7/87, spi and pnc loci: AJ278419; R6, spiR/H locus: AJ278301; strain Sp670, spiR/H locus: AJ278302; spiP peptide gene, strain Hu11: AJ27827; spiP, strain 669: AJ27826; spiP, strain F4: AJ27828; spiP, strain F1: AJ278300; spiP, strain SA) (10760): AJ278299; spiP, strain Sp628: AJ27825; spiP, strain Fi2301r (43351): AJ27824.

3 Results and discussion

3.1 Sequence variation of the peptide SpiP and the adjacent regulatory system in S. pneumoniae clones

The available preliminary genome sequence is derived from a serotype 4 S. pneumoniae isolate KNR7/87. The arrangement of the regulatory system spiR1spiR2spiH, the peptide spiP and putative ABC transporter and accessory domain for processing and export of peptides containing a double glycine-type leader peptide named spiABCD are shown in Fig. 1. The arrangement suggests that the locus represents another peptide-regulated signal transduction system similar to the com loci which are involved in induction of genetic competence: spiP, spiR2 and spiH are homologues of the comCDE operon, and spiABCspiD are related to the comAcomB locus of ABC transporters. The three genes spiABC encode the N-terminal processing domain, the transmembrane domain, and the ATP binding cassette of a peptide exporter, respectively, and spiD is closely related to comB. The region is preceded by a BOX element, a repeat sequence found in intergenic regions of the S. pneumoniae chromosome [16].

Figure 1

Schematic representation of the peptide-regulated spi and the adjacent bacteriocin pnc gene clusters. ORFs are indicated by arrows, putative transcriptional units are indicated by linear arrows on top. Putative ABC transporters and peptides with a double glycine-type leader sequence are shown in black; putative membrane proteins are gray. IS1381 element [24], hatched. P: promoter consensus sequence; T (loops) indicate possible terminator structures. Black rectangles marked BOX: intergenic repetitive element of S. pneumoniae of unknown function [16]; the scale below starts with the A1TG start codon of spiR1.

Preliminary attempts to amplify the peptide gene in S. pneumoniae R6 failed when oligonucleotide primers designed according to the available sequence were used. However, a 4.7-kb region including the entire regulatory locus could be amplified in several genetically distinct S. pneumoniae clones and in one S. mitis isolate, and HinfI restriction analyses documented extensive sequence variation of this locus (not shown). Two strains were subjected to detailed DNA sequence analysis of the three regulatory genes and the peptide: the laboratory strain R6, and strain 670 representing the widespread multiple antibiotic-resistant S. pneumoniae serotype 6B clone [8]. The response regulator SpiR2 and the 112-amino acid putative DNA binding proteins SpiR1 were very similar in the three strains examined (Fig. 2A). In contrast, the 3′-end of the histidine kinase spiH encoding the N-terminal sensor domain differed substantially resulting in 12 and 25% variation at the amino acid level. The variable regions in the sensor kinase were located mainly within the putative membrane spanning domains (Fig. 2B). Also, whereas the 3′-end of the peptide gene encoding the double glycine leader was identical except for one codon in strain 670, the 5′-end encoding the C-terminal mature signaling molecule differed completely, resulting in two distinct mature peptide variants with 28 amino acid residues in contrast to 18 amino acids of the reference strain (Table 1). Sequence comparison of the deduced peptide in a total of 10 strains revealed another 28-residue variant (Table 1). Thus, the signaling molecule and the sensor domain of the receptor kinase occur as distinct functional units, similar to the allelic variants described for the comC/comD pair where activation of the receptor kinase requires the presence of the matching peptide ([17,18] and references therein).

Figure 2

Comparison of the regulatory proteins SpiR1, SpiR2 and SpiH in S. pneumoniae 670 and R6 compared to KNR7/87 clones. A: Deduced peptide sequence of the C-terminal domain of response regulator SpiR2 and of SpiR1. Regions of high similarity between the two proteins are underlined. Only the amino acids that differ from the reference sequences are shown. B: Deduced peptide sequence of the N-terminal domain of the histidine kinase SpiH. Bars indicate predicted transmembrane helices. The conserved histidine occurs at position 254.

View this table:
Table 1

SpiP peptide variants and bacteriocins in S. pneumoniae

The strains represent distinct clones; nc: non-encapsulated. Bacteriocins: strain KNR7/87. The consensus sequence of Gly-Gly leader peptides is given below (cons.). The arrow indicates the processing site. Asterisks indicate conserved sites in the processed SpiP peptides. The Mr and pI are calculated for the processed peptides.

3.2 Peptide-dependent induction of the regulatory locus

To analyze the activity and specificity of the peptide, the 18-amino acid residue of KNR7/87 and the 28-amino acid residue R6 variant were chemically synthesized. They were tested using a promoterless lacZ reporter gene fused the 5′-end of the partial response regulator gene spiR1 in S. pneumoniae R6bgaA::erm. The peptides were added to the growth medium, and β-galactosidase activity determined at various time points. Addition of the KNR7/87 peptide had no effect up to a concentration of 1000 ng ml−1. The R6 peptide, however, stimulated lacZ expression in the R6 derivative demonstrating that the interaction is allele-specific. LacZ activity was induced around 10 ng ml−1 (Fig. 3). With fivefold higher or lower concentrations, no induction occurred. The genes involved in this regulatory system were named spi for S. pneumoniae peptide induction (Fig. 1).

Figure 3

Measurement of β-galactosidase production in R6::pPRO2. Synthetic R6 SpiP peptide was added at time 0. Growth was followed by nephelometry (N, ♦). Aliquots for determination of β-galactosidase activity were taken at the times as indicated. MU: Miller units. Final concentration of SpiP in the growth medium (ng ml−1): 0, △; 2, ◻; 10, ○.

Upstream of the spiR1 promoter an imperfect direct repeat (IDR) was found (Fig. 4). The related ComE response regulator has been shown to bind to an imperfect repeat motif in the comC promoter region [19], and repeats are common in promoter regions of other peptide-regulated genes [20]. In Lactobacillus plantarum, the two response regulators have been shown to bind to the IDRs found within adjacent bacteriocin promoters with different binding properties [21]. It is therefore feasible that similarly, both SpiR1 and SpiR2 bind to the repeat sequences and control adjacent genes.

Figure 4

Sequence and location of the imperfect repeat element within the spi and pnc gene cluster. The four imperfect direct repeats are boxed and numbered. The consensus sequence is given below. The distance in base pairs between the end of the repeat element and the start codon is given on the right; the arrows indicate the orientation of the gene with respect to the repeat.

3.3 A bacteriocin cluster in S. pneumoniae

Three more IDRs could be identified in the KNR7/87 genome, all of which were located adjacent to the spi locus (Fig. 1): one occurred between two divergent operons encoding the ABC exporter system spiABCD and a cluster of small ORFs; and two repeats were present upstream of adjacent transcripts as proposed in Fig. 1. The distance between the IDR and the start codons of putative SpiR-regulated ORFs as well as their orientation with respect to the ATG varied (Fig. 4).

The ORFs encode a reservoir of peptides characteristic for bacteriocin production [1,22]. One single and two tandem ORFs encode bacteriocin homologues of class II with a typical double-glycine leader peptide [23], followed by small hydrophobic peptides resembling immunity proteins generally located in close proximity to bacteriocins; they were named pncApncM for pneumococcal bacteriocins (Fig. 1). Each of the putative bacteriocins had a distinct leader peptide and a glycine-rich mature peptide of 45–66 amino acids in agreement with the general length of antimicrobial peptides between 30 and 60 residues [1], and each contained two cysteine residues (Table 1).

The identification of bacteriocin genes in S. pneumoniae, made possible through genomic analysis, is not surprising, given the fact that they have been observed in many lactic acid bacteria (see [1,22] for review). The spi/pnc locus resembles other class II bacteriocins and their inducing factors described in Lactobacillus sake LTH673, L. plantarum C11, and Carnobacterium piscicola LV17, except that the inducer peptide SpiP is not part of the regulatory transcription unit spiR1spiR2spiH, and that SpiR1 contains only the C-terminal domain of a response regulator, i.e. it cannot be activated by phosphorylation.

Bacteriocin-producing S. pneumoniae may affect not only the growth of pneumococcal strains but other streptococcal species as well. It is tempting to assume that bacteriocin production, i.e. the killing and lysis of other streptococci, could also be beneficial for accessing the gene pool provided by related bacteria.


This work was supported by the Stiftung Rheinland Pfalz für Innovation and the European Community Contract No. BI 04-CT98-0424. The excellent technical assistance of Ulrike Klein is greatly acknowledged. The results were presented at the trap meeting, Kaiserslautern, 12–16 September 1999.


  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].
View Abstract