Characterization of a Novel Fucose-Regulated Promoter (P_fcsK) Suitable for Gene Essentiality and Antibacterial Mode-of-Action Studies in Streptococcus pneumoniae

Abstract

The promoter of the Streptococcus pneumoniae putative fuculose kinase gene (fcsK), the first gene of a novel fucose utilization operon, is induced by fucose and repressed by glucose or sucrose. When the streptococcal polypeptide deformylase (PDF) gene (def1, encoding PDF) was placed under the control of P_fcsK, fucose-dependent growth of the S. pneumoniae (P_fcsK::def1) strain was observed, confirming the essential nature of PDF in this organism. The mode of antibacterial action of actinonin, a known PDF inhibitor, was also confirmed with this strain. The endogenous fuculose kinase promoter is a tightly regulated, titratable promoter which will be useful for target validation and for confirmation of the mode of action of novel antibacterial drugs in S. pneumoniae.

Streptococcus pneumoniae is a widespread human pathogen and a major cause of community-acquired diseases such as pneumonia, otitis media, sinusitis, and meningitis (12). Established antibiotic treatments of pneumococcal infections have become less effective due to the emergence of drug-resistant isolates (34). A genomics-based strategy has been applied in the search for new drug targets to identify inhibitors active against this pathogen (25, 33). Once a lead antimicrobial compound has been discovered, it is fundamentally important to demonstrate that during chemical optimization the antibacterial activity continues to be related to inhibition of the specific target (25). Tightly regulated, titratable promoter systems that are able to modulate the levels of the protein target have proven to be invaluable tools for tracking the mechanism of antibacterial activity of novel inhibitors (2, 37). In addition, inducible promoters have been used in antimicrobial drug discovery for establishing gene essentiality and characterizing the function of essential drug targets (1, 9, 31). For S. pneumoniae only a limited number of regulated promoters have been studied. Heterologous promoter systems derived from nisin and tetracycline genes have been analyzed as tools for regulating gene expression, but their narrow titratable range and high basal levels of expression have compromised their use (1, 2, 7). The streptococcal promoter of the maltose operon has been characterized at the molecular level and shown to be inducible by maltose and repressible by sucrose in S. pneumoniae (23), but its use has been limited by its high basal expression levels. More recently, the raffinose operon has been identified and its promoter has been shown to be regulated, though its application for target validation or mode-of-action analysis has yet to be demonstrated (27). The expression of genes involved in sugar metabolism is known to be a regulated process in many bacterial species. With the availability of genomic sequence data for S. pneumoniae, it is now possible to identify and study many, and perhaps all, putative sugar metabolic genes and their associated promoter sequences.

Identification and bioinformatic analysis of the fucose gene cluster of S. pneumoniae.

In an attempt to identify novel regulatable promoter systems in S. pneumoniae, DNA sequences from three pneumococcal genomes (strains 100993 [GlaxoSmithKline], R6 [15], and type 4 [32]) were examined for homology to known carbohydrate utilization operons of Escherichia coli and Bacillus subtilis. Fourteen putative carbohydrate utilization operons were identified, including those for cellobiose, fructose, fucose, galactose, glucose, lactose, maltose, mannitol, mannose, raffinose, sucrose, and trehalose (data not shown). Several of these operons contained homologs of genes involved in the regulation, uptake, and metabolism of sugars in bacteria. A putative fucose gene cluster containing 11 genes, some of which showed homology to the fucose catabolism genes of E. coli and Haemophilus influenzae (8, 19), was selected for further studies (Fig. 1A). The first gene of the putative operon encodes a protein exhibiting 37% identity to the rhamnulokinase of Salmonella enterica serovar Typhimurium (24) and 22% identity to fuculokinase of H. influenzae (10) and is consequently referred to as fcsK. The gene products encoded by the next two genes in the operon (fcsA and fcsU) showed 41 and 49% amino acid identity to E. coli FucA, a fuculose-1-phosphate aldolase involved in fucose catabolism (38) and H. influenzae FucU, a fucose operon protein (10) distantly related to proteins with oxidoreductase function, respectively. Downstream in the same orientation are four genes (EIIA-fcs, EIIB-fcs, EIIC-fcs, and EIID-fcs) whose products show 34, 35, 28, and 35% homology to mannose- or fructose-specific enzyme II components A, B, C, and D, respectively, of the phosphotransferase phosphoenolpyruvate sugar transport system (26). Two additional genes, designated fcsY and fcsL in this study, encode hypothetical proteins of unknown function (Fig. 1A), although FcsL contains a region of strong similarity to fucose-lectin binding proteins from Anguilla japonica (14). Distal to fcsL is another putative fucose metabolism gene, fcsI, whose product shows 62% identity to H. influenzae FucI, encoding a l-fucose isomerase that has been implicated in fucose catabolism (10). Divergently transcribed to the fcsK operon is a putative regulatory gene encoding FcsR (Fig. 1A), which shows 35% identity to LacR, the Streptococcus mutans lactose repressor (28), suggesting that the fucose operon is subject to negative regulation. DNA sequence analysis revealed putative promoters in the fucose gene cluster located upstream of both fcsK and fcsR (Fig. 1B). The genome organization of the fucose operon is conserved in all three publicly available S. pneumoniae genome sequences (6, 15, 32). However, the role of fucose in pneumococcal metabolism is unclear, since all the S. pneumoniae strains tested were unable to grow either in l- or d-fucose as the sole carbon source in a semidefined medium.

FIG. 1.

Genetic organization of the S. pneumoniae putative fucose (fcs) gene cluster and DNA sequence of its promoter (P_fcsK). (A) The open reading frames and directions of transcription of the 11 genes in the putative fucose cluster are indicated: fcsR (fucose repressor protein), fcsK (fuculose kinase), fcsA (fuculose-1-phosphate aldolase), fcsU (fucose operon protein and putative oxidoreductase), EIIA-fcs to EIID-fcs (enzyme IIA to IID components of the phosphotransferase-phosphoenolpyruvate sugar uptake system), fcsY (hypothetical protein), fcsL (putative fucose-lectin binding protein), and fcsI (l-fucose isomerase). (B) DNA sequence of the intergenic fcsR-fcsK region. The translational start sites of the fcsR and fcsK genes, which are divergently transcribed, are indicated by bold arrows. The open reading frames have been boxed. Putative ribosome binding sites (RBS) and extended −10 and −10 and −35 promoter sequences are highlighted. The 5′-fluorescently tagged primer fcsKpe, used to determine the 5′ end of the fcsK transcript and hence the putative location of the P_fcsK promoter, is shown. For the primer extension reaction, 15 μg of DNase-treated RNA (isolated from S. pneumoniae R6 cells grown in a semidefined medium in the presence of 0.2% glucose and 1% fucose) was denatured together with 5 pmol of fcsKpe primer and reverse transcribed to cDNA. Samples were separated with a Perkin-Elmer ABI 377XL sequencer electrophoresis set, and the size of the primer extension product (196 ± 2 nucleotides) was determined with Perkin-Elmer ABI Prism Genescan Analysis 2.1 software. The transcriptional initiation site (+1) upstream of the fcsK gene, predicted by primer extension analysis, is indicated. A palindromic sequence showing similarity to the catabolite repression element (CRE) sequence of B. subtilis (16), determined with the MAST software program, is boxed.

Transcriptional analysis of P_fcsK by primer extension.

The transcriptional start site of P_fcsK was mapped by primer extension to an adenine residue 24 bp upstream of the fcsK initiation codon (Fig. 1B). Transcription is predicted to start 8 bp downstream of a canonical −10 RNA polymerase binding sequence (TATAAT) which is itself separated by 17 bp from a near-consensus sequence (TTGAAA) for the −35 region (Fig. 1B). A consensus extended −10 promoter (TGTGCTATAAT), which is common in S. pneumoniae (29), was identified upstream of the fucose repressor gene (fcsR), transcribed divergently to fcsK. A palindromic sequence (AGTAAGCGTTCACA) possessing close identity (13 out of 14 bp) to the consensus sequence of the B. subtilis carbon catabolite repression element (TGt/aNANCGNTNa/tCA) was also located in the intergenic region between fcsK and fcsR (Fig. 1B) (16), suggesting that the fucose gene cluster is subject to catabolite repression. The exact role of FcsR and possibly CcpA (carbon catabolite binding protein) (16) in regulating the expression of this operon remains to be determined.

RT-PCR analysis of induction of the fuculose kinase gene (fcsK).

Regulation of the presumptive promoter (P_fcsK) immediately upstream of fcsK was examined by growing wild-type S. pneumoniae R6 under 10 different growth conditions in a semidefined medium (AGCH [a basal medium containing casein hydrolysate, amino acids, vitamins, salts, albumin, and catalase] containing 0.2% [wt/vol] yeast extract [YE] [17]) supplemented with different sugars as carbon source at the time of inoculation. In the presence of glucose, sucrose, lactose, trehalose, fructose, or mannose, added at 1% (wt/vol) concentrations, S. pneumoniae grew to late logarithmic phase (optical density at 650 nm [OD₆₅₀] = 0.6) at approximately the same growth rates (data not shown). Since S. pneumoniae was unable to utilize 1% fucose or galactose for growth and showed a significantly reduced growth rate in the presence of 0.5% raffinose, 0.2% (wt/vol) glucose was added to the medium to support growth in these three cases (27). RNA was prepared from bacteria grown to late logarithmic phase (OD₆₅₀ = 0.6), and the effect of the sugars on the levels of the fcsK transcript was quantified by reverse transcription-PCR (RT-PCR) analysis. Steady-state fcsK mRNA levels under the different growth conditions were compared to those found when the strain was grown in the presence of 0.2% (wt/vol) glucose alone (Fig. 2). Growth in glucose, sucrose, lactose, trehalose, fructose, and mannose had no significant effect on the levels of fcsK transcript (Fig. 2). Transcription of fcsK was induced by fucose and also by the structurally related sugar galactose. Levels of steady-state fcsK mRNA increased 23-fold in the presence of 1% l-fucose and 15-fold following the addition of 1% l-galactose (Fig. 2). Our analysis also indicated that levels of fcsK mRNA are very low in the presence of glucose or sucrose. Although S. pneumoniae does not grow on fucose, the promoter P_fcsK is functional and clearly induced by that sugar.

FIG. 2.

Effects of different sugars on the induction of the fuculose kinase (fcsK) gene measured by quantitative, real-time SybrGreen RT-PCR. S. pneumoniae R6 was grown statically at 37°C in AGCH-YE medium and supplemented with different test sugars. Total RNA was extracted from late-logarithmic-phase-grown S. pneumoniae cells by using the Bio 101 FastRNA kit (Vista, Calif.) following glass bead cell disruption and a hot phenol lysis step (5). DNase-treated RNA was reverse transcribed to cDNA with a First Strand synthesis kit (Invitrogen). Relative levels of bacterial transcripts in each sample were quantified by PCR following SybrGreen dye incorporation (SybrGreen PCR core reagent kit; Applied Biosystems, Perkin-Elmer), and products were detected in real time with the 7700 sequence detection system (Applied Biosystems) as described previously (22, 35, 37). Template primers used in the PCRs are available on request. The quantity of cDNA estimated was normalized to a housekeeping gene, era. Changes in steady-state levels of fcsK mRNA in each sample were expressed relative to the uninduced control (0.2% glucose).

Regulation of P_fcsK::luxAB.

Regulation of the fuculose kinase promoter in S. pneumoniae was further characterized by integrating a P_fcsK::luxAB transcriptional reporter system into the chromosome (Fig. 3A). Since S. pneumoniae is unable to grow on fucose as the sole carbon source, a second sugar substrate (sucrose or glucose) was added to support growth. To analyze regulation of P_fcsK, the S. pneumoniae (P_fcsK::luxAB) reporter strain was grown in media containing 0.3% sucrose and different concentrations of l-fucose added as the inducer at the start of growth (Fig. 3B). There was no difference in the growth rate of the S. pneumoniae reporter under the various conditions, with all cultures reaching a maximum OD₆₅₀ of approximately 0.9 after 8 h (data not shown). Optimal P_fcsK::luxAB expression, as measured by luciferase activity, was observed during late logarithmic phase (OD₆₅₀ = 0.6). The promoter fusion activity was shown to be titratable by varying the concentrations of fucose from 0.01 to 1% in 0.3% sucrose (Fig. 3B). Maximum induction was observed in medium containing >0.4% fucose as the inducing sugar in the presence of 0.3% sucrose (Fig. 3B). Relative to growth carried out in medium lacking fucose, there was an approximately 300-fold increase in luciferase activity following the addition of 0.4% fucose (Fig. 3B). Moreover, repression of P_fcsK by about 25-fold was observed in the presence of increasing concentrations of sucrose (Fig. 3B). A similar titratable effect was demonstrated when glucose was used to replace sucrose as the carbon source (data not shown). Clearly, P_fcsK can be very effectively repressed by sucrose or glucose. Under these test conditions, in which maximum fucose induction is measured in the presence of sucrose (a repressor of P_fcsK), the promoter shows a full dynamic range of approximately 7,500-fold. The measurement of the steady-state transcript levels of luxAB and fcsK by real-time quantitative PCR analysis, from the same reporter strain, following induction with fucose, showed dynamic ranges that correlated with luciferase activity (results not shown). For S. pneumoniae, sucrose-mediated repression of other sugar metabolic pathways has also been reported, though the exact mechanism(s) of this regulation is unknown (18, 23, 27). Transcription analysis with both RT-PCR and reporter gene fusion technology has identified fucose as the primary inducer of the P_fcsK promoter and indicated the extent of its regulation (Fig. 2 and 3).

FIG. 3.

Fucose induction and sucrose repression of P_fcsK in an S. pneumoniae (P_fcsK::luxAB) reporter strain. (A) Construction of a P_fcsK::luxAB transcriptional reporter fusion. A genetic map shows the organization of the P_fcsK::luxAB reporter cassette following integration into the chromosome of S. pneumoniae R6. Details of construction of the luxAB reporter strain in S. pneumoniae are available on request. Briefly, the cassette contains the promoter region of the fuculose kinase gene (P_fcsK) amplified from S. pneumoniae R6 and fused to a promoterless luxAB reporter gene from Vibrio harveyi (13). Transcriptional terminators of two large rRNA operons, TT1 and TT2, were amplified from S. pneumoniae R6 and introduced to flank the reporter fusion and prevent local transcriptional interference. The erythromycin resistance marker (ermAM) of Enterococcus faecalis was amplified from pAMβ1 for selection (21). The cassette was flanked by regions of the β-galactosidase (bgaA) structural gene (36). The construct was integrated into S. pneumoniae R6 at the bgaA locus by transformation (33). Erythromycin-resistant transformants were selected (5 μg of erythromycin/ml), and successful construction of the P_fcsK::luxAB reporter strain in single copy in the chromosome was confirmed by both diagnostic PCR and DNA sequencing. Arrows indicate the directions of transcription of the genes. Lollipop structures represent the transcriptional terminators. (B) Titration range of luciferase activity in the S. pneumoniae (P_fcsK::luxAB) reporter strain following growth in the presence of sucrose and fucose. To study the regulation of P_fcsK in S. pneumoniae, the S. pneumoniae (P_fcsK::luxAB) transcriptional reporter fusion strain was grown to late logarithmic phase (OD₆₅₀ of about 0.6) in AGCH-YE medium containing different concentrations of sucrose and fucose. To measure luciferase activity, bacterial cells (250 μl) were transferred to a microtiter plate and 2 μl of n-decyl aldehyde substrate (Sigma) was added. Light output from the reaction was counted for 2 s with a MicroLumat LB96P luminometer (EG & G Berthold). The relative light units were calculated as the light output per OD₆₅₀ unit per milliliter of culture (3, 13).

Utilization of P_fcsK for essentiality testing.

The gene encoding polypeptide deformylase (PDF) has previously been shown to be essential in E. coli (11), Staphylococcus aureus (37), and S. pneumoniae (2, 20). To investigate the utility of the fucose-regulatable promoter for essentiality testing in S. pneumoniae, a promoter replacement strategy was used to place def1 (encoding PDF) under the control of P_fcsK on the S. pneumoniae chromosome (Fig. 4A) (details of construction available on request). The resulting mutant strain, S. pneumoniae (P_fcsK::def1), was recovered in the presence of fucose. The mutant strain was tested for growth dependency on fucose by monitoring growth in 0.8% sucrose with various amounts of fucose as inducer (Fig. 4B). The higher concentration of sucrose in these experiments supports optimal growth of the control strain and maximizes the extent of repression of the P_fcsK promoter. In the presence of 0.1% fucose, S. pneumoniae (P_fcsK::def1) grew identically to the control wild-type strain S. pneumoniae R6 (data not shown). A decrease in the growth rate of the mutant strain was observed when the levels of fucose in the medium were reduced to 0.05%. Growth was not supported in 0.025% fucose, indicating that PDF levels had fallen below the threshold required for bacterial growth (Fig. 4B). The growth of the control strain S. pneumoniae R6 was not affected by the addition of fucose (results not shown).

FIG. 4.

Fucose-dependent growth of an S. pneumoniae (P_fcsK::def1) regulatable strain. (A) An S. pneumoniae def1 regulatable strain (FD) was constructed by placing def1 under the control of the P_fcsK inducible promoter in the chromosome of S. pneumoniae. Briefly, by using a three-piece PCR strategy of overlapping primers (33), a promoter replacement cassette was constructed containing the P_fcsK promoter, transcriptional terminators (TT1 and TT2) located 5′ of the P_fcsK promoter, and an independent erythromycin resistance marker (ermAM) (21) and flanked by DNA sequences of the gene immediately upstream of def1 and the start of the def1 open reading frame. The construct was integrated into S. pneumoniae R6 at the def1 locus by transformation (33). An erythromycin-resistant mutant of P_fcsK::def1 (FD) was recovered in the presence of added fucose, and both diagnostic PCR and DNA sequencing confirmed its chromosomal organization. (B) The effect of fucose on growth of an S. pneumoniae (P_fcsK::def1) regulatable strain. S. pneumoniae strains were grown statically at 37°C in AGCH-YE medium containing 0.8% (wt/vol) sucrose and l-fucose at 0.1% (•), 0.05% (▪), or 0.025% (▴) (wt/vol). Growth experiments were performed in triplicate in a microtiter plate format with a SpectraMax250 spectrophotometer (Molecular Devices) as described previously (37). (C) Western blot analysis of PDF levels in S. pneumoniae (P_fcsK::def1) FD and R6. S. pneumoniae R6 (lanes 6 and 8) and S. pneumoniae (P_fcsK::def1) FD (lanes 7 and 9) were grown in AGCH-YE medium plus 0.8% (wt/vol) sucrose and fucose at 0.8% (lanes 6 and 7) or 0.1% (lanes 8 and 9) (wt/vol) final concentrations. Mid-logarithmic-phase cultures (OD₆₅₀ of approximately 0.15) were resuspended in sterile distilled water to an equivalent of an OD₆₅₀ of 4 (path length, 1 cm), and total cell lysates were prepared as described previously (30). Ten microliters of total protein samples (lanes 6 to 9) and S. pneumoniae PDF protein standards in 50-, 10-, 2-, 0.4-, and 0.08-ng amounts (lanes 1 to 5, respectively) were loaded onto each lane of a 10% NuPAGE Bis-Tris resolving gel and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (37). Western blotting was performed as described previously (37), and protein samples were probed with rabbit polyclonal antiserum raised against S. pneumoniae PDF (Covance Research Products) (diluted 1/1,000) and anti-rabbit horseradish peroxidase (Sigma) as secondary antibody (diluted 1/10,000).

To demonstrate that the fucose-dependent growth of S. pneumoniae (P_fcsK::def1) was due to a titration of PDF levels in the cell, total cellular protein was prepared from S. pneumoniae R6 and P_fcsK::def1 strains (Fig. 4C) and probed with antibodies raised against S. pneumoniae PDF. PDF protein levels were approximately fivefold lower in the regulatable strain (lanes 7 and 9, Fig. 4C) than in the wild-type strain (lanes 6 and 8, Fig. 4C), even when cells were grown under conditions that support optimal growth (0.8% sucrose and 0.1 or 0.8% fucose [Fig. 4B]). In these studies, the presence of sucrose serves both as the carbon source for growth and as the repressor of P_fcsK. Sucrose-mediated catabolite repression of P_fcsK::def1 may suppress fucose induction, explaining why, even at conditions of maximum induction, the fcsK promoter cannot reach the expression level of the natural def1 promoter. Since there was no difference in growth between wild-type and mutant strains, 20% of wild-type levels of PDF is sufficient (i.e., not rate limiting) for growth under these conditions.

Bioinformatic analysis of the S. pneumoniae genome organization showed that def1 is the first gene in a two-gene operon, which also includes yacO (encoding a putative RNA methyltransferase). Downregulation of yacO is not responsible for the growth defect observed in S. pneumoniae (P_fcsK::def1) at low fucose concentrations because allelic replacement of this gene has shown that it is not essential for in vitro growth in S. pneumoniae (data not shown). The clear dose dependency on fucose for growth of the def1-regulatable strain demonstrates that def1 is indispensable for cell viability in S. pneumoniae.

Application of established S. pneumoniae regulatable systems for the purpose of gene essentiality testing has been limited by relatively high levels of basal expression. For example, regulation of expression of def1 in S. pneumoniae has previously been achieved with the use of a tetracycline promoter system (2); however, no change in cell growth rate was observed in the absence of inducer, indicating that def1 expression cannot be downregulated to the level required for confirmation of essentiality. In contrast, the essential nature of PDF in S. pneumoniae could be demonstrated with the use of P_fcsK.

Clearly, induction by fucose can overcome repression by sucrose, but the full level of induction is not achieved, and as a consequence, strains in which a gene is under the control of P_fcsK generally underproduce the target protein. Given that protein levels can be titrated down with decreasing amounts of inducer, gene essentiality can still be demonstrated.

Application of S. pneumoniae (P_fcsK::def1) for antibiotic mode-of-action studies.

S. pneumoniae (P_fcsK::def1) produces five times less PDF than does the wild-type strain (Fig. 4C) and should therefore be hypersensitive to any PDF inhibitor. To demonstrate the utility of this strain for studying antibiotic mode of action, the MICs of a number of antibiotics were determined for S. pneumoniae (P_fcsK::def1) and R6 following static overnight growth at 37°C in AGCH-YE medium containing 0.8% sucrose and 0.8% fucose. Indeed, the underexpressing def1 strain (MIC = 0.125 to 0.25 μg/ml) showed a 32- to 64-fold increase in sensitivity to actinonin, a potent inhibitor of PDF activity (4), with respect to the wild-type strain (MIC = 8 to 16 μg/ml). In contrast, the sensitivity of the strain to a number of known inhibitors of DNA replication, transcription, translation, cell wall biosynthesis, and fatty acid biosynthesis remained unchanged. These results are consistent with inhibition of PDF as the reason for the antibacterial activity of actinonin and demonstrate the utility of such strains in mode-of-action studies for potential antibacterial agents in S. pneumoniae.

Strains in which expression of the target protein is under regulation are very powerful tools in antibacterial mode-of-action studies because underexpression of the essential protein should lead to a concomitant increase in sensitivity to specific inhibitors. This study is the first case reported in the literature of an endogenous S. pneumoniae promoter whose inducible-repressible characteristics allow its utilization for both essentiality and mode-of-action studies.