Transformation of nonencapsulated Streptococcus pneumoniae during systemic infection

Abstract

Streptococcus pneumoniae (pneumococcus) is a principal cause of bacterial middle ear infections, pneumonia, and meningitis. Capsule-targeted pneumococcal vaccines have likely contributed to increased carriage of nonencapsulated S. pneumoniae (NESp). Some NESp lineages are associated with highly efficient DNA uptake and transformation frequencies. However, NESp strains lack capsule that may increase disease severity. We tested the hypothesis that NESp could acquire capsule during systemic infection and transform into more virulent pneumococci. We reveal that NESp strains MNZ67 and MNZ41 are highly transformable and resistant to multiple antibiotics. Natural transformation of NESp when co-administered with heat-killed encapsulated strain WU2 in a murine model of systemic infection resulted in encapsulation of NESp and increased virulence during bacteremia. Functional capsule production increased the pathogenic potential of MNZ67 by significantly decreasing complement deposition on the bacterial surface. However, capsule acquisition did not further decrease complement deposition on the relatively highly pathogenic strain MNZ41. Whole genome sequencing of select transformants demonstrated that recombination of up to 56.7 kbp length occurred at the capsule locus, along with additional recombination occurring at distal sites harboring virulence-associated genes. These findings indicate NESp can compensate for lack of capsule production and rapidly evolve into more virulent strains.

Figure 1

Survival analysis of C57BL/6 mice during pneumococcal bacteremia. Seven-week-old mice were intraperitoneally infected with indicated pneumococcal strains, and survival was monitored over an 8-day period. Survival rates are shown as percentages (WU2: n = 10, ΔWU2: n = 6, MNZ67: n = 8, MNZ67 + ΔWU2: n = 8, MNZ41: n = 14, MNZ41 + ΔWU2: n = 11). Significantly higher mortality rates were observed in mice infected with WU2 or MNZ67 + ΔWU2 when compared to all other groups. Data are representative of at least two independent experiments. Survival curves were determined to be significantly different (p < 0.0001) by Log-rank (Mantel-Cox) test. ΔWU2 = heat-killed WU2.

Figure 2

Pneumococcal CFUs recovered from the blood of infected mice. Seven- week-old mice were challenged with 10⁸ colony forming units (CFUs). Bacterial loads at 24 h post infection (a) and during the indicated time course (b) were estimated by plating blood mixtures on blood agar plates. At 24 hpi, significantly higher bacterial loads (p < 0.01) were recovered from MNZ41-infected mice compared to MNZ67-infected mice (One-Way ANOVA with Tukey’s post-hoc analysis). From 4–72 hpi, MNZ41-infected mice had significantly higher (p < 0.01) bacterial loads compared to MNZ67-infected mice regardless of co-administration of heat-killed WU2 (Two-Way ANOVA with Tukey’s post-hoc analysis). Data shown represent at least two independent experiments. Dashed line indicates the limit of detection. Error bars represent standard error of the means. **p < 0.01, ***p < 0.001, ****p < 0.0001 vs MNZ67.

Figure 3

Genomic comparison of recombination regions involving capsule locus among donor, transformant, and recipient. For both recombination events (a, b), encapsulated serotype 3 S. pneumoniae strain WU2 served as the donor strain. (a) NESp strain MNZ41 is the recipient and 41S3 is the transformant. (b) NESp strain MNZ67 is the recipient and 67S1 is the transformant. Trans = Transformant, Recip = Recipient. Arrows indicate direction of open reading frame (ORF). Blue arrows indicate either dexB or aliA capsule-flanking genes to demarcate the capsule region. Magenta lines represent contig breakpoints. Black pointers represent recombination breakpoints. Grayscale represents % nucleotide identity between sequences.

Figure 4

Flow cytometry evaluating functional capsule production and deposition of C3b on the surface of viable pneumococci. Pneumococci were incubated with antibodies against Type 3 capsule (a) or normal human serum followed by antibodies against complement component C3b (b). Cells were subsequently stained with Alexa Fluor 488 and subjected to flow cytometry. Percent positive cells were calculated by gating fluorescence higher than background (black line = gate), and averages of percent positive cells ± standard error of the mean are listed to the right of representative histograms in overlays. Mean fluorescent intensities (MFIs) are represented in bar graphs. Data represent at least two independent experiments with 100,000 events collected for each sample. Error bars denote standard error of the mean.