The effects of methionine acquisition and synthesis on Streptococcus pneumoniae growth and virulence

Abstract

Bacterial pathogens need to acquire nutrients from the host, but for many nutrients their importance during infection remain poorly understood. We have investigated the importance of methionine acquisition and synthesis for Streptococcus pneumoniae growth and virulence using strains with gene deletions affecting a putative methionine ABC transporter lipoprotein (Sp_0149, metQ) and/or methionine biosynthesis enzymes (Sp_0585 - Sp_0586, metE and metF). Immunoblot analysis confirmed MetQ was a lipoprotein and present in all S. pneumoniae strains investigated. However, vaccination with MetQ did not prevent fatal S. pneumoniae infection in mice despite stimulating a strong specific IgG response. Tryptophan fluorescence spectroscopy and isothermal titration calorimetry demonstrated that MetQ has both a high affinity and specificity for L-methionine with a K(D) of ∼25 nM, and a ΔmetQ strain had reduced uptake of C(14)-methionine. Growth of the ΔmetQ/ΔmetEF strain was greatly impaired in chemically defined medium containing low concentrations of methionine and in blood but was partially restored by addition of high concentrations of exogenous methionine. Mixed infection models showed no attenuation of the ΔmetQ, ΔmetEF and ΔmetQ/ΔmetEF strains in their ability to colonise the mouse nasopharnyx. In a mouse model of systemic infection although significant infection was established in all mice, there were reduced spleen bacterial CFU after infection with the ΔmetQ/ΔmetEF strain compared to the wild-type strain. These data demonstrate that Sp_0149 encodes a high affinity methionine ABC transporter lipoprotein and that Sp_0585 - Sp_0586 are likely to be required for methionine synthesis. Although Sp_0149 and Sp_0585-Sp_0586 make a contribution towards full virulence, neither was essential for S. pneumoniae survival during infection.

Figure 1. Schematic of the Sp_0148-53 and Sp_0584-0587 loci.

(A1) Structure of the Sp_0148-53 locus. Each arrow represents one gene, with the Sp number stated within the arrow and gene annotation and probable function written below. The number of bp between the stop codon of the upstream gene and the ATG of the downstream gene is given above the arrows in the corresponding gap. (A2) Representation of the structure of the Sp_0148-53 locus in the ΔmetQ strain; the erythromycin resistance cassette (ery^r) is shaded grey. (B1) Structure of the Sp_0584-0587 locus, with the Sp number stated and gene annotation when available written below. The number of bp between the stop codon of the upstream gene and the ATG of the downstream gene is given above the arrows in the corresponding gap. (B2) Structure of the Sp_0584-0587 locus in the ΔmetEF mutant strain, showing replacement of metEFgenes with an in-frame copy of cat (shaded grey).

Figure 2. Immunoblots of S. pneumoniae using polyclonal anti-MetQ antibodies.

(A) Probing of whole cell lysates, lipid and aqueous phases of Triton X-114 extracts of wild-type, Δlsp, and ΔmetQ 0100993 strains, and the recombinant MetQ protein (missing the N terminal signal sequence) with anti-MetQ antibodies. Equal numbers of bacteria were used for each strain, and the approximate sizes of the bands are displayed on the right. (B) Immunoblots of representative strains for common S. pneumoniae serotypes probed with anti-MetQ antibodies, showing the presence of a similar band in all strains.

Figure 3. Analysis of L-methionine binding to MetQ.

(A) Using tryptophan fluorescence spectroscopy: MetQ (0.05 µM) in 50 mM potassium phosphate pH 8 was excited at 280 nm at 37°C. Example titration of the MetQ signal change at 342 nm with increasing concentrations of L-methionine. (B) Binding isotherms for the interaction of MetQ with L-methionine. The top panel display heat changes upon injection of ligand and the low panel the integrated hears of injection (▪) and the best fit (solid line) to a single site binding model. (C) Radioactive uptake of ¹⁴C-methionine expressed as nmol/mg protein over time in minutes by the wild-type (circles, 50 µM of methionine added) and ΔmetQ (open triangles 50 µM of methionine added, filled triangles 750 µM of methionine added) S. pneumoniae strains.

Figure 4. Investigation of MetQ as a vaccine candidate.

(A) Mouse sera antibody titres to recombinant His-tagged MetQ after intraperitoneal immunisation with His₆-MetQ. Target protein (His₆-MetQ or the negative control SBP His₆-LivJ) and IgG class are labelled along the X axis, and results were obtained using serum from His₆-MetQ vaccinated mice with the exception of the filled triangles which represent results for serum from control mice (labelled alum). Filled symbols represent titres for individual mice for the total IgG titre, open circles for IgG1, and open diamonds for IgG2A titres. Bars represent median values. (B) Development of fatal disease in mice vaccinated with His₆-MetQ (square symbols) and control (given alum and PBS alone, diamonds) mice after IP inoculation of 10⁴ CFU of D39 (n = 20). There were no statistical differences in survival (log rank test). (C) Phagocytosis of representative S. pneumoniae vaccine serotypes. Key: bacteria opsonised with: grey bars, HBSS; solid bars, serum from alum immunised mice; dark grey bars, polyclonal anti-MetQ sera from MetQ vaccinated mice.

Figure 5. Gene expression of Sp_0149 during infection.

(A) Ethidium stained agarose gel showing equal amplification efficiency for the primer pairs used for PCR of psaA and metQ when using S. pneumoniae 0100993 genomic DNA as the template. (B) Ethidium stained agarose gel of cDNA products generated by RT-PCR after 26 and 30 cycles using S. pneumoniae 0100993 total RNA extracted from the blood of infected mice and primers for 16S rRNA (internal control), psaA (positive control) and metQ. (C) Relative mRNA concentrations of selected genes of S. pneumoniae WCH43 (serotype 4) in various in vivo niches at 72 h post-intranasal infection of mice. Transcript abundance for each gene was obtained by microarray analysis, and normalized against those obtained for the 16S rRNA control. Quantitative fold differences for each transcript were determined as a ratio of its abundance to that of the 16S rRNA control. Data are means ± SEM for each gene transcript from three replicate challenge experiments.

Figure 6. Growth of the wild-type, ΔmetQ, ΔmetEF and ΔmetEF/ΔmetQ S. pneumoniae strains in nutrient rich and semi defined media.

Growth measured using OD₅₈₀ of the wild-type (squares), ΔmetQ (inverted triangles), ΔmetEF (triangles) and ΔmetEF/ΔmetQ (diamonds) S. pneumoniae strains grown in THY (A), C+Y with no added methionine (B), C+Y with 50 µM methionine (C), or C+Y with 750 µM methionine (D). Each data point is the mean log₂ OD₅₈₀ for three samples.

Figure 7. Growth of the wild-type, ΔmetQ, ΔmetEF and ΔmetEF/ΔmetQ S. pneumoniae strains in human blood or in human blood supplemented with amino acids.

Data is presented as the mean (SD) fold change in the bacterial CFU per ml at 2 h (clear columns) and at 4 h (grey columns) in human blood (A) or at 4 h (clear columns) and 6 h (grey columns) in human blood supplemented with amino acids (B). P values for the differences between the wild-type and the mutant strains at 4 h (A) and between the ΔmetEF/ΔmetQ strain with and without added methionine at 4 h or 6 h (B) were obtained using ANOVA and post hoc tests, *P<0.05, ** p<0.01. Log¹⁰ bacterial CFU per ml results for each condition are given in figures above the corresponding bar in the graphs.

Figure 8. Virulence of the ΔmetQ, ΔmetEF and ΔmetEF/ΔmetQ strains in mouse models of infection.

(A) CIs for mixed infections with the ΔmetQ, ΔmetEF and ΔmetEF/ΔmetQ and wild-type strains for bacteria recovered from the nasal washes 5 days after colonisation. Each point represents results for a single mouse, and bars median CIs for the group. (B) Log₁₀ ml⁻¹ bacterial CFU recovered from lungs, BALF and spleens of mice 24 hours after IN inoculation with 1×10⁷ CFU of either the wild-type or ΔmetQ strain. There were no statistical differences between the strains for any organ site. (C) Log₁₀ ml⁻¹ bacterial CFU recovered from blood 24 hours after IP inoculation with 3×10³ CFU of either the wild-type or ΔmetEF/ΔmetQ strain. P = 0.0317, Mann Whitney U test.