Parallel Evolution in Streptococcus pneumoniae Biofilms

Abstract

Streptococcus pneumoniae is a commensal human pathogen and the causative agent of various invasive and noninvasive diseases. Carriage of the pneumococcus in the nasopharynx is thought to be mediated by biofilm formation, an environment where isogenic populations frequently give rise to morphological colony variants, including small colony variant (SCV) phenotypes. We employed metabolic characterization and whole-genome sequencing of biofilm-derived S. pneumoniae serotype 22F pneumococcal SCVs to investigate diversification during biofilm formation. Phenotypic profiling revealed that SCVs exhibit reduced growth rates, reduced capsule expression, altered metabolic profiles, and increased biofilm formation compared to the ancestral strain. Whole-genome sequencing of 12 SCVs from independent biofilm experiments revealed that all SCVs studied had mutations within the DNA-directed RNA polymerase delta subunit (RpoE). Mutations included four large-scale deletions ranging from 51 to 264 bp, one insertion resulting in a coding frameshift, and seven nonsense single-nucleotide substitutions that result in a truncated gene product. This work links mutations in the rpoE gene to SCV formation and enhanced biofilm development in S. pneumoniae and therefore may have important implications for colonization, carriage, and persistence of the organism. Furthermore, recurrent mutation of the pneumococcal rpoE gene presents an unprecedented level of parallel evolution in pneumococcal biofilm development.

Keywords: Streptococcus pneumoniae; biofilm; genomic diversification; parallel evolution.

Fig. 1.—

CFU enumeration of biofilm-derived pneumococcal variants. Cells were harvested from triplicate pneumococcal biofilms from two independent experiments at time points 1, 3, 6, and 9 days and plated onto CBA to assess for changes in colony morphology. Variants were defined as follows: SCVs (<0.5 mm), MCVs (0.5–1 mm), and TCVs (>1 mm). Data represent the mean CFU counts from the two biofilm experiment. Error bars represent standard error. Images of colonies were taken on CBA under ×6 objective on a Leica dissection Stereomicroscope. Scale bar = 1 mm.

Fig. 2.—

Colony quantification of serotype 22F biofilm-derived colony variants. Diameter values of the three distinct biofilm-derived colony variant populations harvested from pneumococcal biofilms quantified using ImageJ analysis software. A total of nine colony variants from triplicate biofilms were measured. Numbers signify mean values.

Fig. 3.—

Biofilm formation of the biofilm-derived colony variants. Three-day-old biofilms of (A) the WT ancestral strain, (B) SCV, (C) MCV, and (D) TCV were visualized on a Leica TCS SP5 confocal laser scanning microscope using BacLight live/dead stain. Green cells indicate live cells and red cells indicate dead cells. Z-Scans were performed every 0.3 µm on each field of view. White bars represent 25 µm.

Fig. 4.—

COMSTAT analysis of biofilm-derived colony variants. Biofilm formation of the biofilm-derived variants quantified using COMSTAT 1 and the program Matlab (R2012a) (Heydorn et al. 2000) for triplicate 3-day-old biofilms of each variant type, grown in MatTek culture plates under static conditions. Graph depicts (A) the mean average thickness and maximum thickness, (B) mean biomass, and (C) mean surface area. Error bars represent 95% confidence intervals.

Fig. 5.—

Mean growth rate of biofilm-derived SCV phenotype. The growth rate was assessed for SCVs from the exponential growth phase (between 4 and 8 h of growth) from triplicate growth experiments. SCV rate was calculated using pooled data from all 12 sequenced SCVs. A two-sample t-test was used to compare SCV rate to the WT rate. Numbers signify mean values. Error bars represent 95% confidence intervals.

Fig. 6.—

Capsule quantification of the SCV phenotype. Capsule quantification was determined by staining the pneumococcal capsule with Stains-all solution. The absorbance was measured at optical density 640 nm and subtracted from the negative control. Capsule quantification was calculated using pooled data from all 12 sequenced SCVs; graph depicts mean absorbance. Error bars represent 95% confidence intervals. Graph represents pooled data from all 12 sequenced SCVs.