Promoter Identification and Transcription Analysis of Penicillin-Binding Protein Genes in Streptococcus pneumoniae R6.

Penicillin-binding proteins (PBPs) are membrane-associated enzymes, which are involved in the last two steps of peptidoglycan biosynthesis, and some of them are key players in cell division. Furthermore, they are targets of β-lactams, the most widely used antibiotics. Nevertheless, very little is known about the expression and regulation of PBP genes. Using transcriptional mapping, we now determined the promoter regions of PBP genes from the laboratory strain Streptococcus pneumoniae R6 and examined the expression profile of these six promoters. The extended -10 region is highly conserved and complies with a σ(A)-type promoter consensus sequence. In contrast, the -35 region is poorly conserved, indicating the possibility for differential PBP regulation. All PBP promoters were constitutively expressed and highly active during the exponential and early stationary growth phase. However, the individual expression of PBP promoters varied approximately fourfold, with pbp1a being the highest and pbp3 the lowest. Furthermore, the deletion of one nucleotide in the spacer region of the PBP3 promoter reduced pbp3 expression ∼10-fold. The addition of cefotaxime above the minimal inhibitory concentration (MIC) did not affect PBP expression in the penicillin-sensitive R6 strain. No evidence for regulation of S. pneumoniae PBP genes was obtained.

Introduction

Penicillin-binding proteins (PBPs) are modular membrane-bound enzymes catalyzing the final steps of bacterial cell wall synthesis. They are the targets of β-lactam antibiotics and play important roles in the division process. Streptococcus pneumoniae contains six PBPs, which are classified with respect to their molecular weight, domain structure, and enzymatic activities into three classes (for reviews, see Goffin and Ghuysen[1]; Sauvage et al.[2]; Zapun et al.[3]). All PBPs contain a DD-peptidase domain. S. pneumoniae PBP1a, PBP1b, and PBP2a are class A high-molecular weight (HMW) PBPs. They are bifunctional enzymes, since they polymerize the glycan chains by their N-terminal glycosyltransferase domain and crosslink the peptides by their DD-transpeptidase domain. The members of class B HMW PBPs (PBP2x and PBP2b) are monofunctional DD-transpeptidases and contain an N-terminal domain of unknown function. In addition, PBP2x contains a C-terminal domain consisting of two PASTA (PBP- and serine/threonine kinase-associated) domains.[4-6] The topology of all HMW PBPs consists of a cytoplasmic tail and a transmembrane anchor followed by two or three surface-exposed domains.[7] Finally, PBP3 (DacA) is a class C, low-molecular weight (LMW) PBP with DD-carboxypeptidase activity, hydrolyzing the C-terminal D-alanine moiety from the pentapeptides in the peptidoglycan (PG) chain.[8-10]

The genes encoding class A HMW PBPs can be deleted individually, demonstrating that none of them is essential for growth under laboratory conditions.[11,12] It is possible to isolate double mutants pbp1b pbp2a and pbp1a pbp1b, but pbp1a pbp2a double mutants are not viable.[12,11] Both class B PBPs, PBP2x and PBP2b, are essential in S. pneumoniae.[13-15]

S. pneumoniae grows and divides by alternating cycles of peripheral and septal PG synthesis (reviewed in Massidda et al.[16]). The genes pbp2x and pbp2b are located in the dcw (division and cell wall) cluster,[16,17] already indicating a role in the division process. The coordinated function of each PBP during the S. pneumoniae cell cycle remains largely unknown.[18] All HMW PBPs localize at mid cell,[19] the central growth zone, where new cell wall material is incorporated. Recently, the essential roles of PBP2x and PBP2b in cell division were confirmed.[14-16,20,21] Depletion of pbp2b results in the formation of rounded and chained cells, indicating that PBP2b is essential for peripheral PG synthesis.[14,21] The depletion of pbp2x results in lemon-shaped or elongated cells often with pointed ends, showing that PBP2x is responsible for septal PG synthesis.[14,15,21] In contrast, LMW PBP3 does not localize at the division sites like the HMW PBPs, but is distributed over the entire cell surface[22,23] and at mid cell in some dividing cells.[23] Cells lacking PBP3 show heterogeneity in cell size and shape and display defects in septum placement[23-25] indicating an important role of PBP3 during cellular growth in general.

Beta-lactam antibiotics mimic the terminal D-Ala-D-Ala moiety of the PG pentapeptide stem and, therefore, are recognized as PBP suicide substrates.[26] PBPs are inhibited by β-lactams by forming a covalent PBP-β-lactam complex through the active site serine, which is enzymatically inactive. For most beta-lactams, this complex is very stable corresponding up to several generation times; kinetic parameters describing PBP-beta-lactam interactions can be found in Zapun et al.[3] The consequence is inhibition of bacterial growth and cell lysis. Mutations in PBPs of β-lactam-resistant strains reduce the protein affinity to the antibiotic, while leaving the enzymatic function apparently unaffected. Such mutated enzymes do not interact with β-lactams over a wide antibiotic concentration range and, therefore, PG synthesis and cell growth can continue. Mutations in S. pneumoniae PBP2x and PBP2b result in low-level resistance and additional alterations in PBP1a are required for high resistance levels. Occasionally, alterations in PBP2a, 1b, and 3 have been described in high-level-resistant strains (for review, see Hakenbeck et al.[7]). The interaction between PBP and β-lactam antibiotic is specific for each protein and for each antibiotic (for review, see Sauvage and Terrak[27]).[28] PBP2b does not interact with cefotaxime or other β-lactams with similar side chains.[29] In contrast to penicillins, which induce rapid lysis in S. pneumoniae, cells stop growing after cefotaxime treatment without significant lysis for several hours and are killed at a much lower rate. That is because cefotaxime induces a tolerant response.[29] PBP2b mutations can be selected with penicillin,[30] whereas cefotaxime selects for PBP2x mutations.[31-33]

Despite the important roles of PBPs in bacterial cell wall synthesis, cell division processes and development of β-lactam resistance, very little is known about the expression and regulation of PBP genes. To get insights into the expression and potential regulatory mechanisms, we identified the promoters of PBP genes and studied their expression during growth in a penicillin-sensitive strain in the absence and presence of cefotaxime below, at, and above the minimal inhibitory concentration (MIC) values. We show that all PBP promoters are constitutively expressed and the presence of cefotaxime does not influence their transcriptions. In addition, we show that a deletion in the spacer region between the −35 and −10 elements in the PBP3 promoter has a very strong effect on the promoter activity.

Materials and Methods

Bacterial strains, plasmids, and growth conditions

All S. pneumoniae strains are derivatives of S. pneumoniae R6 strain.[34] The strains and plasmids used in this study are described in Table 1. S. pneumoniae strains were grown at 37°C without aeration in C medium[35] supplemented with 0.1% yeast extract (C + Y medium), in brain heart infusion broth (BHI, Broth), or on D-agar[36] plates containing 3% defibrinated sheep blood. Growth in liquid culture was monitored by nephelometry and is given in nephelo units [N].

1.

Bacterial Strains and Plasmids Used in This Study

	Relevant genotype or description[a]	Source or reference
Strains
R6	Unencapsulated, nonvirulent descendent of D39	[34]
RP200	R6, bgaA::tetM-lacZ, TetR	[42]
RKL44	R6, bgaA::tetM-PvegW-lacZ, TetR	[42]
801	Derivative of R6, containing hexB mutation	[43]
KP01	R6, bgaA::tetM-Ppbp1a-lacZ, TetR	This study
KP02	R6, bgaA::tetM-Ppbp1b-lacZ, TetR	This study
KP03	R6, bgaA::tetM-Ppbp2a-lacZ, TetR	This study
KP04	R6, bgaA::tetM-Ppbp2b-lacZ, TetR	This study
KP05	R6, bgaA::tetM-Ppbp2x-lacZ, TetR	This study
KP06	R6, bgaA::tetM-Ppbp3-lacZ, TetR	This study
KP09	R6, bgaA::tetM-Ppbp3-R801-lacZ, TetR	This study
Plasmids
pPP2	Integrative promoter probe plasmid (bgaA::tetM-lacZ), AmpR, TetR	[42]
pPP2vegW	pPP2 derivative, carries PvegW-lacZ fusion	[42]
pPP21a	pPP2 derivative, carries Ppbp1a-lacZ fusion	This study
pPP21b	pPP2 derivative, carries Ppbp1b-lacZ fusion	This study
pPP22a	pPP2 derivative, carries Ppbp2a-lacZ fusion	This study
pPP22b	pPP2 derivative, carries Ppbp2b-lacZ fusion	This study
pPP22x	pPP2 derivative, carries Ppbp2x-lacZ fusion	This study
pPP23	pPP2 derivative, carries Ppbp3-lacZ fusion	This study
pPP23M	pPP2 derivative, carries Ppbp3-R801-lacZ fusion	This study

Antibiotic resistance marker.

Amp, ampicillin; Tet, tetracycline.

Escherichia coli DH5α was used as a host for cloning and propagation of plasmids. E. coli was grown at 37°C either in LB medium with aeration or on LB agar plates.[37] Growth of E. coli was followed by measuring the optical density (OD) at 600 nm.

Transformation and DNA manipulations

Transformation of S. pneumoniae was performed according to a published procedure.[38] Transformants were selected on D-agar supplemented with 3% defibrinated sheep blood and 3 μg/ml tetracycline. E. coli DH5α was transformed according to Hanahan[39] and transformants were selected in the presence of 100 μg/ml ampicillin.

All DNA techniques were performed using standard methods.[37] Plasmid DNA was purified using the NucleoBond AX-100 (MACHEREY-NAGEL) Kit. PCR reactions were performed using GoldStar Taq DNA polymerase (Eurogentec) or high-fidelity iProof DNA polymerase (Bio-Rad Laboratories) according to the manufacturer's instructions. PCR products and DNA recovered after restriction with endonuclease were purified using a JetQuick DNA Purification Kit (Genomed). DNA-modifying enzymes were purchased from New England BioLabs or Invitrogen and used as described by the manufacturer. DNA oligonucleotides used in this study are listed in Table 2 and were obtained from Eurofins, MWG or Operon. RNA oligonucleotides were obtained from Biomers GmbH.

2.

Oligonucleotides Used in This Study

Oligonucleotides	Sequence (5′-3′)
For 5′-RACE amplification
pbp1a_RACE_1	AAGCTAATGCTCAGATACTTGATTAGG
pbp1a_RACE_2	ATTCCTTGTTGGGCAAGGAC
pbp1b_RACE_1	CACGAATCACCGCCTTGGGTACTACAC
pbp1b_RACE_2	AAGTGCGCAACAAATCACTC
pbp2a_RACE_1	GGTAATGGTAGAGCCACCACCTGAACG
pbp2a_RACE_2	CGGCCATAGTTAATCCCGTC
pbp2x_RACE_1	TTCAGCCGGCGATTTCCGATTTTTGG
pbp2x_RACE_2	CTCACGCTGGTCCAATTGAG
pbp2b_RACE_1	CTGATGCTCACATAAGTCAGTAACTTT
pbp2b_RACE_2	ACGCGTAAAGGAAACAACCTGCTTTAAC
pbp3_RACE_1	ATTGACAACAGTGGCATCCTGAATTCC
pbp3_RACE_2	TCTCAGCTAGGGCAATAGCG
RACE-PCR_5′	GATATGCGCGAATTCCTG
5′ RACE-adapter[a]	GAUAUGCGCGAAUUCCUGUAGAACGAACACUAGAAGAAA
For promoter probe cloning[b]
pbp1a_ppf	AGATGCATGCGATCCTTATCACAAAATAAACCC
pbp1a_ppr	GCGGGATCCACATCACTTTATTATACCATAATTTG
pbp1b_ppf	ACATGCATGCCAAAAATGTGCATGTGTTTTTTCTCC
pbp1b_ppr	GCGGGATCCACTTTATCTATTATACCACAAAAGGG
pbp2a_ppf	ACATGCATGCTGAATTGCATACAGATATTGTAAC
pbp2a_ppr	GCGGGATCCTTCATGCGTTTATTTTATCATCTTC
pbp2x_ppf	ACATGCATGCGTCAAGTATTTTTGACATATTTTTTG
pbp2x_ppr	GCGGGATCCATATCTTATCTATTTTACCACAAATC
pbp3_ppf	ACATGCATGCTGGTCCAAAAAAATAATTTATTTCTAC
pbp3_ppr	GCGGGATCCCATAGTAGTCTTATTCTATCATAAAG
pbp2b_ppf	ACATGCATGCGGGAATAGTCACACTTCATTATAAC
pbp2b_ppr	GCGGGATCCATCTTACTCTTAATTGTACCACAATTTG

RNA oligonucleotide obtained from Biomers GmbH.

SphI and BamHI restriction sites are underlined.

RACE, rapid amplification of cDNA ends.

RNA isolation

To isolate total RNA from S. pneumoniae, the cells were grown in C + Y medium to a density of N = 70, pelleted by 8.000 rpm at 4°C, and frozen in liquid nitrogen. For all subsequent solutions, diethylpyrocarbonate-treated water (DEPC; Ambion) was used. RNA was extracted as described previously.[40] Residual DNA was digested by the addition of 5 units (U) DNase (New England BioLabs) and incubation for 15 min at 37°C. The RNA was purified further using the RNeasy Midi Kit (QIAGEN) according to the manufacturer's instructions.

Mapping of transcriptional start sites

The transcriptional start point of each PBP transcript was determined by rapid amplification of cDNA ends (5′-RACE) as described previously[41,42] using the same RNA adapter and gene-specific oligonucleotides (Table 2). Briefly, 15 μg of total cellular RNA was incubated with and without 25 U Tobacco Acid Pyrophosphatase (TAP; Epicentre Biotechnologies) in the buffer supplied by the manufacturer at 37°C for 60 min in the presence of 20 U SUPERaseIN RNase inhibitor (Ambion). Subsequently, the reaction mixture was extracted with phenol:chloroform (1:1) and precipitated with ethanol. RNA pellets were resuspended in 55 μl DEPC-treated water (Ambion), mixed with 500 pmol RNA adapter (5′ RACE-Adapter; Table 2) (Biomers GmbH) and heat-denatured at 95°C for 5 min. Ligation of the adapter was carried out at 17°C overnight with 100 U T4 RNA ligase (New England BioLabs) in the supplied buffer and in the presence of 80 U SUPERaseIN (Ambion) in a total reaction volume of 100 μl. The reaction mixture was extracted with phenol:chloroform (1:1), ethanol precipitated, and the pellet was resuspended in 25 μl DEPC-treated water. Subsequently, 5 μl of ligated RNA were reverse transcribed with a gene-specific oligonucleotide (Table 2) using the First-Strand cDNA Synthesis Kit for RT-PCR (AMV; Roche) according to the instructions of the manufacturer in a volume of 20 μl. Then, 2 μl of cDNA were amplified by PCR using the nested gene-specific oligonucleotides (Table 2) and the RACE-PCR_5′ oligonucleotide, which is complementary to the RNA adapter sequence. The PCR was performed using the GoldStar Taq DNA polymerase (Eurogentec) for 40 cycles at an annealing temperature of 50°C. Subsequently, PCR fragments were analyzed on a 2% agarose gel and sequenced.

The sequenced PCR fragments of pbp1b, pbp2a, and pbp2b showed in electropherograms a weak additional sequence in the region of the RNA adapter sequence, which overlaps with high-quality sequence and suggests the presence of the multiple sequences. To exclude the possibility that two transcription starts are present, the PCR fragments were ligated into the pGEM-T-Easy (Promega) vector and 15 colonies were analyzed by PCR followed sequencing of the respective PCR products (see text for details).

Construction of plasmids and strains

Transcription was monitored using the promoter probe plasmid pPP2.[42] Promoter regions were amplified by PCR from S. pneumoniae R6 genomic DNA using oligonucleotide pairs, which were named according to the PBP genes as listed in Table 2 (for example, pbp1a_ppf and pbp1a_ppr for the pbp1a promoter). The PCR products were cleaved with SphI and BamHI and inserted into the same sites in pPP2, generating the plasmids pPP21a, etc. Cloning was performed using the E. coli DH5α strain. After sequencing the insertions, plasmids were transformed into S. pneumoniae R6. The promoter probe plasmid cannot replicate in S. pneumoniae and promoter fusions were integrated into the endogenous β-galactosidase gene bgaA leading to its inactivation.[42] Correct integration occurred by double crossover and was confirmed by PCR and DNA sequencing. The resulting S. pneumoniae strains, harboring different promoter-lacZ fusions, were designated KP01–KP06 (Table 1).

The reporter plasmid pPP23M was constructed following mainly the same procedure, except that the promoter fragment of pbp3 was amplified by PCR using genomic S. pneumoniae 801 DNA.[43] The PCR fragment was cleaved with SphI and BamHI and ligated into the same sites into the pPP2 vector, generating the plasmid pPP23M. The resulting S. pneumoniae strain, harboring P-lacZ fusion, was designated KP09.

Determination of β-galactosidase activity

Determination of β-galactosidase activity in cell extracts of S. pneumoniae strains carrying E. coli lacZ gene fusions was performed as described previously.[44,45] The strains were grown in C + Y or BHI media and the β-galactosidase activity was measured at various time points during the exponential growth up to the beginning of the stationary phase. The sample size was adjusted to contain the equivalent of 2 ml cells at N = 90. Specific β-galactosidase activities are expressed in nmoles of nitrophenol released per min per mg of protein. Protein concentrations were determined using the standard method of Bradford.[46] Experiments were carried out in triplicates. Student's t-test was applied to check the significance of the results.

Prediction of promoters and terminators

The promoters were predicted by two consensus hexamers: −10 (5′-TATAAT-3′) and −35 (5′-TTGACA-3′) located upstream of the transcriptional start sites. Rho-independent transcription terminators were identified using the Mfold program,[47] which can be accessed through the The mfold Web Server. The consensus sequence covering the promoter region was designed using the web-based application WebLogo.[48,49]

Results

Genetic organization of the PBP genes in S. pneumoniae R6 genome

The genes encoding PBPs in S. pneumoniae are localized at six distinct loci (Fig. 1) scattered throughout the genome. We searched for putative promoter regions and putative transcriptional terminators to define the genetic organization of S. pneumoniae R6 genomic regions containing the PBP genes.

Organization of genomic regions containing Streptococcus pneumoniae R6 PBP genes. The coding regions and their directions of transcription are indicated by arrows and drawn in scale as indicated by the bar representing the size of 1 kb. PBP genes are shown in red, genes organized in operons with PBP genes are white, and flanking genes are imaged in gray. BOX and RUP elements are depicted in blue. Putative promoters are indicated by small black arrows and putative transcriptional terminators are shown as stem-loop structures. Spr numbers represent S. pneumoniae R6 genes. GI number of the S. pneumoniae R6 chromosome sequence in the NCBI Nucleotide database is GI:15902044; pbp1a (pbpA) identified in the NCBI Gene database (GeneID:934791), pbp1b (GeneID:934893), pbp2a (GeneID:933569), pbp2b (GeneID:933948), pbp2x (pbpX, GeneID:934744), and pbp3 (dacA, GeneID:934315). GI, GenInfo identifier; PBP, penicillin-binding protein.

The PBP1a gene is apparently cotranscribed together with the upstream RecU gene. The transcription starts at a promoter upstream of recU and ends at a strong transcriptional terminator behind pbp1a. The two genes overlap by four nucleotides (nt). This transcriptional unit was also identified by genomic tiling arrays.[50] The RecU gene essential in S. pneumoniae[51] encodes the Holliday junction resolvase RecU, which in Staphylococcus aureus is required for chromosome segregation and DNA damage repair.[52]

The transcript of pbp1b is monocistronic. Although the region between pbp1b and the downstream-located gene spr1908 encoding a hypothetical protein is only 64 nt long, it contains a putative terminator followed by the spr1908 promoter.

The genetic organization of the pbp2a locus strongly suggests that pbp2a is part of a putative operon, including rpmG, secE, and nusG encoding a putative competence-specific global transcription modulator, the SecE subunit of the preprotein translocase, and a transcription termination/antitermination factor. A promoter has been identified upstream of pbp2a gene and only a very weak terminator was found downstream of pbp2a. The presence of an additional promoter downstream of pbp2a gene indicates that rpmG-secE-nusG can be independently transcribed. Coexpression of rpmG-secE was identified in the S. pneumoniae TIGR4 strain by genomic tiling arrays.[50]

Furthermore, the PBP2b gene forms an operon with recR located downstream. A promoter was predicted in front of pbp2b and a putative transcriptional terminator downstream of the RecR gene encodes a protein, which may play a role in DNA repair. No information is available regarding the essentiality of RecR in S. pneumoniae.

The genetic organization of the pbp2x locus indicates that the four genes constitute an operon, which was described as cell wall gene cluster by Massidda et al.[17] A promoter has been identified in front of the MraW gene (encoding a S-adenosyl-methyltransferase) and a terminator downstream of mraY (encoding a phospho-N-acetylmuramoyl-pentapeptide-transferase) followed by a RUPA repeat element.[53] PBP2x is the third gene in this operon downstream of ftsL (encoding a cell division protein of unknown function), in which all genes are essential and their products are involved in PG biosynthesis.[16,51,54]

Finally, the transcript of pbp3 appears to be monocistronic ending at a transcriptional terminator just before a BOXABC element.[55] A promoter structure with an extended −10 region was found upstream of the PBP3 gene.

Determination of the transcription initiation site

To determine the transcriptional start sites of the six pbp promoters, 5′-RACEs were performed using oligonucleotides located in the downstream genes of the predicted promoters (Fig. 1 and Table 2). RNA of the R6 strain was treated with and without the TAP enzyme, which converts the 5′-triphosphate of RNA to a monophosphate and thereby enables ligation of the RNA to an adapter oligonucleotide. The adaptor allows specific amplification of the 5′-end sequence subsequent to reverse transcription. Finally, the PCR fragments were sequenced and the start points were determined for the six transcripts that include the PBP genes (Fig. 2). The DNA sequence electropherograms of six PCR products showed high-quality sequences and thereby clearly identified the transcription start sites (Fig. 2). The transcriptional start of pbp1a is located 620 nt upstream from the pbp1a start codon and 27 nt upstream of the recU start codon, which is the first gene in operon (Fig. 1). For pbp1b, the transcription start was mapped to the adenine residue located 13 nt upstream of the ATG start codon. The pbp2a transcriptional start overlaps with the putative translational start codon, which implies that the mRNA is leaderless; no ribosomal binding site (RBS) was detected upstream of the start codon. The transcriptional start of pbp2x was mapped 1,286 nt upstream from the pbp2x start codon and is located only 3 nt upstream of the mraW start codon. Similarly, the transcriptional start of pbp3 was mapped as the adenine residue located 3 nt upstream of the pbp3 start codon. The spacer between the transcriptional and putative translational initiation sites for both, pbp2x (mraW) and pbp3, is also too short to accommodate an RBS. Finally, for pbp2b, the transcripts initiate at an adenine residue located 17 nt from the first putative start codon of pbp2b.

Determination of transcriptional start sites of PBP genes using 5′-RACE. Electropherograms from the DNA sequencing reactions of the 5′-RACE PCR products are shown. The sequences of the RNA adaptors are shown as gray bars. The nucleotides complementary to the 5′ ends of mRNA are shown and the directions of the transcription are indicated by arrows. The first nucleotide represents the transcription initiation site. RACE, rapid amplification of cDNA ends.

In addition to high-quality sequence profiles, the electropherograms of three PCR products pbp1b, pbp2a, and pbp2b (Fig. 2, right) show weak sequences, in particular, noticeable in the area of the sequence complementary to the RNA adaptor. An explanation for these results could be that mRNA molecules are degraded by endonuclease. To exclude the possibility that two transcription starts are present, the PCR fragments were cloned into the pGEM-T-Easy vector and 15 colonies were analyzed by PCR and sequencing. For pbp1b, one second transcription start, and for pbp2a, two additional transcription starts were identified. In all cases, transcription starts were located within the coding region with no obvious −10 region and −35 consensus regions close by, therefore, they were not considered in subsequent experiments. In case of pbp2b, no additional transcriptional starts were found.

Consensus sequence of S. pneumoniae promoters

Based on the mapping results described above, we aligned the promoter regions of PBPs loci and analyzed the sequences (Fig. 3A). At a distance 6–8 nt upstream of the transcriptional start site, almost all sequences exhibit an extended −10 promoter element (5′-TRTGNTATAAT-3′), which was originally identified in Bacillus subtilis,[56] and which is common in S. pneumoniae.[57] The extended −10 element 5′-TRTG-3′ is located 1 nt upstream of the −10 hexamer, which consists of the 6 nt TATAAT and is absolutely essential for transcription. The 5′-TG-3′ motif of the extended −10 promoters is positioned at −15/−14 with respect to the transcriptional start site. The extended −10 element is highly conserved in all promoters of the PBP loci (Fig. 3B). In contrast, the −35 element is not well conserved and is difficult to define. Only the two promoters upstream of pbp1a and pbp1b contain, respectively, four and five matches to consensus −35 hexamer. However, in both cases, the length of the spacer between the promoter −10 and −35 elements is not optimal and contains 19 or 15 nt, respectively. For E. coli, it has been shown that the most frequent spacer region length is 17 nt.[58] The promoter of pbp2a contains four matches to the consensus −35 hexamer sequence and has a 17 nt long spacer. The other three promoters displayed a poor match to the −35 hexamer with two or three matches only. Figure 3B depicts a DNA sequence logo generated from the alignment of pbp promoters from Fig. 3A. The best conserved promoter element is clearly the −10 sequence.

Alignment of putative S. pneumoniae R6 PBP promoters. (A) The alignment shows the conserved sequences located upstream of the six PBP loci. The −35 regions marked in orange; the extended −10 region is shown in green within the gray area. Experimentally determined transcriptional start points are shown in blue. The first translation start codons (ATG) are shown in red. (B) The WebLogo[48,49] representation of the position weight matrices derived from the sequences depicted in (A) is shown.

No other regulatory motifs common to some or all promoter regions of the PBP loci were found, and thus it is not likely that one regulatory mechanism directs the expression of all PBP genes.

PBP promoter activity in the sensitive R6 strain

To evaluate pbp promoter activity, the integrative promoter probe system pPP2 was used.[42] PBP promoter fragments were cloned upstream of the promoterless E. coli β-galactosidase LacZ gene. In the next step, the reporter constructs were stably integrated into the S. pneumoniae R6 genome by homologous recombination at the bgaA locus encoding the endogenous β-galactosidase. Upon integration by double crossover, most of the BgaA gene was deleted and the background β-galactosidase level was eliminated. As controls, the RP200 and RKL44 strains were used, which carry a promoterless-lacZ and a P-lacZ fusion, respectively. The vegW promoter served as a nonregulated control for regulation studies in S. pneumoniae R6 derivatives. P is a derivative of P and was constructed by reducing the spacing between the −10 and the −35 region from 18 to 17 nt, thereby deleting a G.[42] Subsequently, β-galactosidase activities were determined throughout the growth of the strains in C + Y medium. As shown in Fig. 4, all PBPs promoters are constitutively expressed and highly active during the exponential growth and early stationary growth phase as well, similar to the constitutive promoter P, which showed a β-galactosidase activity of 200 U throughout; the promoterless control construct gave values below 1 U (data not shown). All data were validated by the Student's t-test implying no growth phase-specific regulation of PBP expression in the sensitive R6 strain. Interestingly, the individual expression activities of the PBP promoters varied. Lowest promoter activities were observed for PBP3 and PBP1b and the strongest promoter was PBP1a, leading to the following hierarchy of the PBP promoter activities PBP3 < PBP1b < PBP2x = PBP2a < PBP2b < PBP1a (Fig. 4). Promoter activities were also determined in cells grown in BHI medium since this medium is easy to prepare and the measurements of β-galactosidase activities provide stable results. Consistently, the same relative expression patterns were observed, but the levels of expression were ranging between 2.1- and 2.8-fold higher than in C + Y medium (data not shown). The expression of P also increased to the same degree (400 U). A higher expression of a variety of promoters in cells grown in BHI medium has been reported[45]; the reason for this phenomenon is unclear.

Activity of the PBP promoters in S. pneumoniae R6 strain during growth in C + Y medium. Growth curves and β-galactosidase activities of one representative experiment are shown. Growth was monitored by nephelometry (N, nephelo units). β-galactosidase activities were determined in 30 min intervals and are given in nmol nitrophenol produced per min and mg of protein.

Response of promoter activities to cefotaxime

Beta-lactams interact with the DD-peptidase domain of PBPs resulting in growth inhibition of S. pneumoniae.[2,7,28] Cefotaxime was chosen since this compound induces a tolerant response and the cells do not lyse for many hours.[29] In contrast, most other beta-lactams induce a rapid lytic response in this organism and consequently a different experimental setup would be required to prevent cellular lysis during the sampling period. However, even during nonlytic conditions beta-lactam-treated cells shed membrane vesicles and precursors of cell wall polymers into the medium,[29,59] and the interpretation of results under such artefact prone conditions causes severe problems.

We tested whether the presence of cefotaxime in BHI medium affects the expression of PBP genes. Bacterial growth was inhibited already at a concentration of 0.01 μg/ml (0.5 × MIC) cefotaxime and the cells stopped growing at a cell density of N = 30 (Fig. 5A). The expression of PBP genes was determined at 30, 60, and 150 min after the addition of cefotaxime at a concentration of 0.04 μg/ml (2 × MIC). No significant effect on the expression of PBP genes was observed according to Student's t-test. The expression of pbp2x is shown as an example in Fig. 5B, and expression of all PBP genes at 2 × MIC cefotaxime are shown in Supplementary Fig. S1 (Supplementary Data are available online at www.liebertpub.com/mdr). The addition of cefotaxime at a concentration of 0.4 μg/ml (20 × MIC) did not affect the expression level of pbp2x (Fig. 5B), but R6 cells start to lyse two hours later. After 30 min growth in the presence of 0.4 μg/ml, only PBP2b could be labeled with BOCILLIN™FL, documenting that all other PBPs were acylated (not shown).

Growth and pbp2x promoter activity in response to cefotaxime treatment. (A) Growth of S. pneumoniae R6 in BHI medium was followed by nephelometry (N). Cefotaxime was added to the exponentially growing cultures at the concentrations (μg/ml) at the time point as indicated by the arrow. (B) Activity of the pbp2x promoter at different time points after the addition of cefotaxime. 0 indicates the time immediately before the cefotaxime addition. β-galactosidase activities were measured after 30, 90, and 150 min and are given in nmol nitrophenol produced per min and mg of protein. Black bars: control, no cefotaxime; gray bars: 0.04 μg/ml cefotaxime; white bars: 0.4 μg/ml cefotaxime. The results are mean value ± SD of three independent experiments. SD, standard deviation.

A mutation in the promoter region affects the amount of PBP3

While investigating the physiological role of PBP3 and its effect on cefotaxime resistance, Selakovitch-Chenu et al. used the laboratory strain 801, an R6 derivative, which produces a reduced amount of PBP3.[43] They described that this strain carries a deletion of 1 nt in the upstream sequence of the PBP3 gene.[60] We sequenced the promoter region of the pbp3 gene from mutant 801 to identify the position of this mutation within the promoter region defined in the present study. The deletion of 1 nt is located in the space region between the −35 and the extended −10 elements (Fig. 6A). It is well established that the length of the spacer region in E. coli is critical.[61] Usually, the −35 sequence is located 17 nt upstream of the −10 element.[58] In case of the 801 strain, the length of the PBP3 promoter spacer region is 16 nt. Therefore, we decided to evaluate the promoter activity and cloned this promoter in the plasmid pPP2. The resulting plasmid pPP3M was transformed into the S. pneumoniae R6 and the reporter construct was integrated into the genome by homologous recombination. Subsequently, β-galactosidase activities were measured during growth at four time points (Fig. 6B). The promoter activity of the mutated PBP3 was reduced 10-fold in comparison to wild-type promoter indicating that the deletion of 1 nt in the space region greatly affects the promoter strength, in agreement with the lower PBP3 protein production in strain 801.

Effect of the mutation in the PBP3 promoter of S. pneumoniae 801 on PBP3 expression. (A) Comparison of the pbp3 promoter region between S. pneumoniae R6 and the 801 mutant. The −35 (orange) and extended −10 promoter (green within the gray area) regions are boldfaced. The transcriptional start point is shown in blue and the translation start codon of pbp3 is indicated in red. The deletion of one nucleotide in 801 is highlighted by the red box. (B) β-galactosidase activities expressed from the promoters pbp3R6 (gray) and pbp3R801 (black) in KP06 and KP09 strains. Strains were grown in C + Y medium. β-galactosidase activities were determined at four different time points and are given in nmol nitrophenol produced per min and mg of protein. The results are mean value ± SD of three independent experiments.

A lower amount of PBP3 has been described also in another laboratory strain I41R.[62] This strain differs from R6 and its derivatives by expressing the DpnII phenotype.[63] Sequencing of the pbp3 promoter region showed that S. pneumoniae I41R carries the same mutation as strain 801 described above.

Discussion

The expression and potential regulatory mechanisms of PBPs are poorly understood in S. pneumoniae. We have now investigated the promoter sequences of the six PBP genes from S. pneumoniae strain R6 in detail and examined the PBP expression pattern under the particular conditions.

The DNA sequences 150 nt upstream and 100 nt downstream (data not shown) of the transcriptional start sites did not reveal any common regulatory motif. Therefore, it is unlikely that one common mechanism controls the expression of all PBP genes in S. pneumoniae. All PBP promoters contained a highly conserved extended −10 promoter element, but the −35 promoter element varied considerably. These results are similar to those obtained by Haenni et al.,[64] who investigated PBP promoters in S. gordonii. This organism also belongs to the Mitis group of viridans streptococci, but is only distantly related to S. pneumoniae.[65] Extended −10 promoter elements occur frequently in Gram-positive bacteria[66] and are also common in S. pneumoniae.[57] In Gram-negative bacteria, such as E. coli, the extended −10 element has a 5′-TG-3′ motif at promoter positions −14 and −15 corresponding to 1 nt upstream of the −10 hexamer.[67] In contrast, Gram-positive bacteria contain a highly conserved extended sequence with the consensus 5′-TRTG-3′, which is also known as the −16 region[68,69] (Fig. 3). The 5′-TRTG-3′ motif was found in five PBP promoters; only the pbp2a promoter contained a 5′-TG-3′ motif. The full −10 extension is at least four times more common in S. pneumoniae than in E. coli.[57] The impact of a full −10 extension compared with the 5′-TG-3′ motif is not well understood. For E. coli, it has been demonstrated that the 5′-TG-3′ motif further stabilizes the interactions between the RNA polymerase and DNA.[70]

Mitchell et al. showed that in E. coli, the extended −10 promoters tend to have longer spacer, have fewer matches to the consensus −35 hexamer, and contain short runs of T residues in the spacer region.[58] We note that three PBP promoters contain 2/6 or 3/6 matches to the consensus −35 hexamer (Fig. 3), whereas the pbp2a promoter shows the highest match (4 out of 6 nt). In all four promoters, the −35 sequence is located 17 nt upstream of the −10 element, which is the most frequent spacer distance in E. coli.[61] Two promoters, pbp1a and pbp1b, have four and five matches, respectively, to the consensus sequence, but in both cases the length of the spacer region is not optimal. Thus, in all PBP promoters the −35 element is not well conserved. In this context, it is important to note that this region plays an important role in the pbp3 promoter activity, although it has only 2 nt matches to the consensus −35 hexamer. The deletion of only 1 nt in the spacer region of PBP3 promoter reduces pbp3 expression ∼10-fold (Fig. 6); whether this is due to a shortened spacer between the −35 and −10 element, or whether the deletion of one thymidine is responsible for this effect remains to be clarified. In contrast, Sabelnikov et al. showed that the S. pneumoniae dpnM promoter,[57] which contains an extended −10 promoter element, can function without a −35 element. Taken together, the general importance of −35 elements in extended −10 promoters in S. pneumoniae bacteria remains unclear.

PBP expression analysis in S. pneumoniae R6 during different growth phases and in two different media did not reveal any specific PBP regulation, since all six pbp promoters were constitutively expressed in both media. However, the individual expression rate of the six PBP promoters differed by approximately a factor of four. It could be possible that different 5′ regions of the transcribed mRNA molecules influence the expression of the reporter LacZ. However, in the publication where the lacZ-fusion system was described, the authors also performed qPCR experiments and did not see any deviations when different promoter regions were tested in the lacZ assay.[44] Previously, it was demonstrated that in E. coli, the expression of the LMW PBP6a was 2- to 10-fold higher in stationary phase than in exponentially growing cells.[71] We did not see an alteration of the expression in early stationary growth phase. The expression pattern during late stationary growth was not followed. Nevertheless, no change in expression pattern was observed between exponential growth and early stationary phase, independent on the medium used in the experiments (C + Y vs. BHI medium) (see Fig. 4 for C + Y). This is in agreement with data obtained from microarray analysis assessing the responses of S. pneumoniae to penicillin.[72]

Haenni et al. showed that in the presence of low concentrations of penicillin, pbp2a of S. gordonii displayed increased expression, and in a penicillin-resistant mutant, penicillin treatment resulted in increased expression of pbp1a as well.[64] The authors suggested that an increased transglycosylase activity could stabilize a poorly crosslinked PG, and in doing so, promote bacterial survival at borderline penicillin concentrations. In contrast, the expression of all six S. pneumoniae PBPs was not affected upon addition of 0.04 μg/ml cefotaxime (2 × MIC) in BHI medium. This concentration already inhibits bacterial growth (Fig. 5), but does not affect the localization profiles of PBP2x tagged with GFP.[15]

Although we could not detect any significant changes in the expression pattern of the PBP promoters in this study, it cannot be ruled out that other regulatory mechanisms exist that control the PG biosynthesis machineries or components thereof during cellular growth directly or indirectly. Further analysis of PBP expression and biochemical data during treatment with different beta-lactam antibiotics and in resistant strains of S. pneumoniae and other, more closely related streptococci of the Mitis group, will help to shed more light on potential regulatory mechanisms of PBPs in these bacteria.