Pyruvate formate lyase is required for pneumococcal fermentative metabolism and virulence

Abstract

Knowledge of the in vivo physiology and metabolism of Streptococcus pneumoniae is limited, even though pneumococci rely on efficient acquisition and metabolism of the host nutrients for growth and survival. Because the nutrient-limited, hypoxic host tissues favor mixed-acid fermentation, we studied the role of the pneumococcal pyruvate formate lyase (PFL), a key enzyme in mixed-acid fermentation, which is activated posttranslationally by PFL-activating enzyme (PFL-AE). Mutations were introduced to two putative pfl genes, SPD0235 and SPD0420, and two putative pflA genes, SPD0229 and SPD1774. End-product analysis showed that there was no formate, the main end product of the reaction catalyzed by PFL, produced by mutants defective in SPD0420 and SPD1774, indicating that SPD0420 codes for PFL and SPD1774 for putative PFL-AE. Expression of SPD0420 was elevated in galactose-containing medium in anaerobiosis compared to growth in glucose, and the mutation of SPD0420 resulted in the upregulation of fba and pyk, encoding, respectively, fructose 1,6-bisphosphate aldolase and pyruvate kinase, under the same conditions. In addition, an altered fatty acid composition was detected in SPD0420 and SPD1774 mutants. Mice infected intranasally with the SPD0420 and SPD1774 mutants survived significantly longer than the wild type-infected cohort, and bacteremia developed later in the mutant cohort than in the wild type-infected group. Furthermore, the numbers of CFU of the SPD0420 mutant were lower in the nasopharynx and the lungs after intranasal infection, and fewer numbers of mutant CFU than of wild-type CFU were recovered from blood specimens after intravenous infection. The results demonstrate that there is a direct link between pneumococcal fermentative metabolism and virulence.

FIG. 1.

Schematic representation of reactions downstream of pyruvate in lactic acid bacteria. After entering bacteria, galactose is converted to pyruvate, which is then further catabolized by homolactic or mixed-acid fermentation pathways. iPFL, inactive PFL; PDH, pyruvate dehydrogenase; POX, pyruvate oxidase; ADH, alcohol dehydrogenase; ACK, acetate kinase; PTA, phosphotransacetylase.

FIG. 2.

Schematic representation of genomic regions containing putative pfl genes (SPD0235 and SPD0420) (B and C) and pflA genes (SPD0229 and SPD1774) (A and D). The chromosome is represented with a thin solid line, and genes are shown with a block arrow. The chevrons in SPD0229, SPD0235, SPD0420, and SPD01774 represent the approximate positions of insertion (see the text for details), and the direction of chevrons represents the orientation of the spectinomycin cassette. The diagram is not drawn to scale.

FIG. 3.

In vitro growth characteristics of pneumococcal strains in CDM supplemented with glucose (A) and galactose (B). D39, SPD0420M, SPD0420Comp, and SPD1774M are represented by solid; short, dashed; long, dashed; or dotted lines, respectively. Black lines are for aerobiosis, and gray lines are for anaerobiosis. Error bars show the standard error of the mean for three independent experiments.

FIG. 4.

Fermentation end product analysis for strains D39 (A), SPD0420M (B), SPD0420Comp (C), and SPD1774M (D). The pneumococcal strains were grown aerobically or anaerobically in CDM containing glucose or galactose. Glu, glucose; Gal, galactose. Each metabolite concentration was measured in duplicate with supernatants of three independently grown cultures.

FIG. 5.

Growth of D39 (•), SPD0420M (▪), SPD1774M (▴), and SPD0420Comp (▾) in blood specimens (A), nasopharynx (B), and lung (C) after intranasal infection. Each point is the mean of data from five mice. Error bars show the standard error of the mean.

FIG. 6.

Time course of bacterial growth in blood specimens from mice infected intravenously with D39 (•) or SPD0420M (▪). Each point is the mean of data from 6 to 10 mice, except 72 h for SPD0420M, which is from 3 mice. Error bars show the standard error of the mean.