Fig 1.
Morphogenesis of cardiac microlesions.
(A) Representative transmission electron microscopy (TEM) images of cardiac sections (magnification: 2,500X) from BALB/cJ mice infected with S. pneumoniae strain TIGR4 between 24 and 42 hours post-infection (n = 12). Panels A.1-7 depict the morphogenesis of cardiac microlesions beginning as pneumococci-containing microscopic vesicles within the myocardium. Hydropic degeneration (black bold arrows) and mitochondrial damage as evidenced by swelling (white arrows) adjacent to cardiac microlesions are evident. Images that are the most representative of what occurs during individual microlesion development are shown. The images do not necessarily depict the overall course of infection in a mouse which is mixed with different sized microlesions at late timepoints. (B) Representative high power (60,000X) TEM images of pneumococci within microlesions show heterogeneous capsule expression: (B.1) pneumococci within smallest vesicles surrounded by myocardium; (B.2) pneumococci at the periphery of larger microlesions; (B.3) pneumococci within the center of a larger microlesion. (C) Representative TEM image of TIGR4 within a 48-hour old static biofilm (n = 3) grown in a 6-well plate (3,000X). Inset, Representative high power (60,000X) TEM image of biofilm-pneumococci.
Fig 2.
S. pneumoniae within cardiac microlesions exhibit biofilm properties.
(A) In vitro antimicrobial tolerance assays for survival of paired blood-isolated pneumococci (BIP) and heart-isolated pneumococci (HIP) from the same mouse (n = 6). Tolerance to penicillin and erythromycin killing at the designated concentration was tested. Statistical analysis was performed using Mann-Whitney test. (B) Representative high magnification immunofluorescent microscopy images of cardiac microlesions showing presence of biofilm extracellular matrix components: nucleic acids (stained with DAPI [DAPI], blue), capsule (stained with anti-serotype 4 capsule antibody [CPS], red), and mouse extracellular DNA (stained with anti-mH2A histone antibody [mH2A], green). A minimum of 5 stained heart sections were examined. (C) Representative tile-stitched image of whole heart sections from TIGR4 infected mice (n = 5). The cardiac sections (stained with DAPI, blue) were probed for TIGR4 (red), using serotype 4 capsule polysaccharide antisera, and for exposed galactose residues (green), using fluorescein labeled Erythrina crystagalli lectin, within the heart. Inset, Representative tile-stitched image of an uninfected control heart (n = 3). (D) Static biofilm-forming ability of TIGR4 (T4) and T4 ΔspxB was assessed in 48-hour 6-well plate model (n = 3 biological replicates, each with 2 technical replicates). Biofilm biomass was measured using crystal violet staining. Statistical analysis was performed using Student’s t-test. (E) Average number of cardiac microlesions detected per cardiac section in T4 and T4 ΔspxB infected mice post infection. For each time point, cardiac sections from at least 5 mice were examined. Averages were calculated by enumerating the number of microlesions in 3 non-adjacent cardiac sections. Statistical analysis was performed using Mann-Whitney test. P value: * ≤ 0.05, ** ≤ 0.01, **** ≤ 0.0001; data are represented as mean ± SEM.
Fig 3.
Heart isolated pneumococci exhibit tissue tropism.
(A) Percentage of pneumococci in the transparent phenotype recovered from the blood (blue) and heart (red) of infected mice 18, 30, and 42 hours post-infection (n = 5 mice per time point). The TIGR4 parent wildtype strain used as infection inoculum was composed of 84% transparent pneumococci. Multiple group analysis was performed using non-parametric One-way ANOVA (Kruskal-Wallis Test) with Dunn’s multiple comparison test. Comparison of percent transparent colonies of pneumococci isolated from blood or heart at each time points was performed using Mann-Whitney test. (B) Schematic representation of the serial infection model used to assess the invasive abilities of paired BIP (blue) and HIP (red) samples. Three mice were infected with TIGR4 to give 3-paired HIP and BIP samples. Each of these HIP and BIP samples were then used to infect 3 more mice per sample to have a total of 9 HIP-infected mice and 9 BIP-infected mice (3 samples x 3 mice each), i.e. n = 18 total mice. The schematic denotes only one replicate of a 3-part experiment. (C) Pneumococcal titers in the blood of mice infected with TIGR4 BIP or HIP over time. The figure denotes blood titers of mice such that each line denotes the mean bacterial titers in blood of the 3 mice that received the same HIP or BIP sample (total n = 9 HIP and BIP infected mice). Paired samples are denoted by line pattern. Statistical analysis was done between groups at each time point using Mann-Whitney test. No statistical significance was observed between the BIP and HIP blood titers at any time point. (D) Representative tile-stitched images of whole frozen heart sections from mice infected with BIP and HIP. Cardiac sections (stained with DAPI, blue) were probed for TIGR4 using serotype 4 capsule polysaccharide antisera (CPS, green). (E) Average number of cardiac microlesions per mouse heart section following challenge with BIP (n = 7) or HIP (n = 9). Averages were calculated by enumerating the number of microlesions in 3 non-adjacent cardiac sections. Statistical analysis was performed using Mann-Whitney test. (F) Adhesion and invasion of HIP (n = 4) compared to BIP (n = 4) to HL-1 mouse atrial cardiomyocytes in vitro. Values are expressed as fold-increase in HIP relative to BIP. Experiments were done using 4 sets of paired HIP and BIP samples collected from 4 individual mice (i.e. 4 biological replicates). Each sample pair was tested against each other using 3 technical replicates on each cell line. The average of each set of technical replicates, was used to create the figure panel and for statistical analysis. Statistical analysis was performed using Mann-Whitney test. P value: * ≤ 0.05, ** ≤ 0.01; data are represented as mean ± SEM.
Fig 4.
Comparative gene expression analysis of HIP, BIP and pneumococci from in vitro biofilm and planktonic pneumococci.
(A) Dot plot representation of the whole genome transcriptomic profile for blood- isolated pneumococci, BIP (blue) and heart-isolated pneumococci, HIP (red) showing the average normalized number of RNA-seq reads identified, i.e. gene expression levels mapping to each TIGR4 gene within the blood and the heart. Y-axis denotes normalized expression levels (RPKMs) whereas X-axis denotes location of the genes on the TIGR4 chromosome. Established virulence determinants with expression levels >1000 RNA-seq reads (i.e. corresponding to top 10% of genes with highest expression levels) for BIP and HIP are indicated. (B) Dot plot representation of the differential gene expression profile for BIP and HIP spanning the TIGR4 genome. The fold changes are depicted as Log2(HIP/BIP). Y-axis denotes log fold changes in gene expression levels whereas X-axis denotes location of the genes on the TIGR4 chromosome. Important differentially up-regulated pneumococcal genes for BIP and HIP are indicated in blue and red respectively. Genes clustered near the X-axis are consistently expressed. (C) Curve plot representation of gene expression levels for genes encoding designated pneumococcal virulence determinants in the BIP, HIP, in vitro biofilm-, and in vitro planktonic- TIGR4 samples. Y-axis denotes normalized expression levels (i.e. RPKMs) whereas X-axis denotes individual nucleotide location (nt coordinates) on the TIGR4 chromosome. Two pooled BIP samples (5 mice per sample), three HIP samples, three in vitro biofilms and three in vitro planktonic pneumococci samples were tested.
Fig 5.
Pneumococcal transcriptome in the heart is distinct from the blood and in vitro conditions.
(A) Two-dimensional (PC1-PC2) principal component analysis on transcriptomes of BIP, HIP, in vitro biofilm-, and in vitro planktonic TIGR4 samples. PC1 separates the in vivo conditions from the in vitro conditions. PC2 separates biofilm conditions from planktonic conditions. B) Whole transcriptome comparisons for differentially expressed genes under the above conditions presented as circular plots (Circos software, see methods). Ideogram (outer rim): Each quadrant denotes one condition with 1,969 genes ranked by their expression. Red denotes highly expressed genes, gray indicates intermediate expressed genes, and green indicates low expressed genes for each condition. Genes with counts of 0 reads across all samples were excluded. Scatter Plot (inner rim): The scatter plot illustrates the Log10 (Count Per Million-mapped- reads) values for each gene in each condition ranging from -2 (inner) to 5 (outer) in steps of 0.5. Red dots are highly expressed genes; black dots are intermediate expressed genes and green dots are low expressed genes. Links (interior): Red arcs link genes with high expression in both conditions. Green arcs connect genes with low expression in both conditions. Blue arcs link genes with low expression in one condition and high expression in another condition. A higher density of connectivity for links of same color indicates transcriptomic similarities within the pneumococcal populations isolated from the tested conditions. (C) Curve plot representation of gene expression levels for Regions of Diversity RD2, RD6 and RD12 in the BIP, HIP, in vitro biofilm-, and in vitro planktonic- TIGR4 samples. Y-axis denotes normalized expression levels (i.e. RPKMs) whereas X-axis denotes individual nucleotide location (nt coordinates) on the TIGR4 chromosome. RD corresponds to Regions of Diversity as determined by Tettelin and Hollingshead [24]. The Tpr/Phr peptide quorum sensing-signaling cassettes within the RDs are indicated as determined by Hoover et al [46]. Two pooled BIP samples (5 mice per sample), three HIP samples, three in vitro biofilms and three in vitro planktonic pneumococci samples were tested.
Fig 6.
Deletion of RD12 abrogates TIGR4 biofilm formation in vitro, results in a pro-inflammogenic phenotype in vivo, and elicits a stronger immune response from macrophages.
(A) Static biofilm-forming ability of isogenic RD2 (T4ΩRD2) and RD12 (T4ΩRD12) deficient mutants relative to TIGR4 (T4) was assessed in a 48-hour 6-well polystyrene plate model (n = 7 experiments). Biofilm biomass was measured using crystal violet staining. Statistical analysis was performed using Student’s t-test. Shown is a representative image of crystal violet stained biofilms. (B) Representative Z-stacked images of TIGR4, T4ΩRD2, and T4ΩRD12 biofilms grown on coverslips for 24 hours. Biofilms were examined for viability using Live/Dead staining. Whole (live and dead) bacterial biomass stain green whereas non-viable bacteria stain red (n = 3 experiments). (C) Pneumococcal titers in the blood of mice (n = 5 per group) infected with TIGR4 (T4), T4ΩRD2 (ΩRD2), or T4ΩRD12 (ΩRD12) 30 hours post-infection. Mann-Whitney test comparing the mutant strain titers to the wildtype TIGR4 titers was performed. (D) Representative immunofluorescent stained images of cardiac sections from mice infected with TIGR4 (T4), T4ΩRD2 (ΩRD2), or T4ΩRD12 (ΩRD12) 30 hours post infection (n = 5 mice per group). Cardiac sections were stained using serotype 4 capsule polysaccharide antisera (green) and DAPI (blue). (E) Absolute numbers of infiltrated neutrophils in hearts of mice (n = 5 per group) infected with TIGR4, and T4ΩRD12 (ΩRD12) 30-hours post-infection. Neutrophils were identified as Gr-1+CD11b+Ly-6G+F4/80- cells. Statistical analysis was performed using student’s t-test. (F) In vitro antimicrobial tolerance assays for survival of paired blood-isolated pneumococci (BIP) and heart-isolated pneumococci (HIP) from T4ΩRD12 infected mice (n = 5). Tolerance to penicillin and erythromycin killing at the designated concentration was tested. Statistical analysis was performed using Mann-Whitney test. No statistically significant differences were observed. (G) LDH release cytotoxicity assay of J774A.1 macrophages challenged with equal biomass of TIGR4 (T4) and T4ΩRD12 (ΩRD12) as determined at 0, 1, 2, 4 hours post-infection (n = 3 biological replicates, each with 3 technical replicates). Statistical analysis for comparisons of cytotoxicity at each time point was performed using Mann-Whitney test. (H) TNFα and CXCL2 production by J774A.1 macrophages and HL-1 cardiomyocytes following 4-hour exposure to an equal biomass of planktonic TIGR4 (T4), biofilm TIGR4 (T4), or planktonic T4ΩRD12 (ΩRD12) (n = 3 biological replicates, each with 3 technical replicates). Statistical analysis was performed using non-parametric One-way ANOVA (Kruskal-Wallis Test). P value: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001; data are represented as mean ± SEM.
Fig 7.
Heart invaded biofilm pneumococci subvert host immune response by releasing pneumolysin and rapidly kill cardiac macrophages.
(A) Western blots for pneumolysin levels in equal biomass of whole cell lysates (pellets) and supernatants of planktonic- wildtype TIGR4 (n = 3), biofilm- wildtype TIGR4 (n = 3), and planktonic T4ΩRD12 (ΩRD12) (n = 2). An isogenic pneumolysin deficient TIGR4 strain (T4 Δply) was tested as the negative control. Normalized densitometric quantification of pneumolysin levels in the supernatant is provided. Statistical analysis was performed by comparison supernatant pneumolysin levels from planktonic (PK)- wildtype TIGR4 (n = 3), and planktonic T4ΩRD12 (ΩRD12) (n = 2) to biofilm (BF)- wildtype TIGR4 (n = 3) using Welch’s t-test. (B) LDH release cytotoxicity assay of J774A.1 macrophages challenged with equal biomass of planktonic-, biofilm- TIGR4 (T4), planktonic-, biofilm- T4 Δply and planktonic-, biofilm- T4 Δply complemented with exogenous recombinant pneumolysin (rPLY, 0.3μg/mL) as determined at 0, 1, 2, 4 hours post-infection (n = 3 biological replicates, each with 3 technical replicates). Statistical analysis was performed using ordinary one-way ANOVA. (C) TNFα production by J774A.1 macrophages at designated time points following exposure to an equal biomass of planktonic-, biofilm- TIGR4 (T4), planktonic-, biofilm- T4 Δply and planktonic-, biofilm- T4 Δply complemented with exogenous recombinant pneumolysin (rPLY, 0.3μg/mL) as determined at 0, 1, 2, 4 hours post-infection (n = 3 biological replicates, each with 3 technical replicates). Statistical analysis was performed using ordinary one-way ANOVA. (D) Representative transmission electron microscopy (TEM) image of cardiac sections (magnification: 2,500X) from BALB/cJ mice infected with T4 Δply 30 hours post-infection (n = 3). (E) Representative high magnification immunofluorescent microscopy images of cardiac microlesions from uninfected-, passively immunized (αPly)- and naïve- mice infected with HIP or BIP 30 hours post infection, showing presence of: capsule (stained with anti-serotype 4 capsule antibody [CPS], red), cardiac macrophages (stained using anti-Mac-3 antibody [Mac-3], green), and infiltrated neutrophils (stained with anti-Ly-6G antibody [Ly-6G], green). A minimum of 4 stained heart sections were examined. (F) Absolute numbers of infiltrated neutrophils in hearts of uninfected-, passively immunized (αPly)- and naïve- mice infected with HIP or BIP 30 hours post infection. Neutrophils were identified as Gr-1+CD11b+Ly-6G+F4/80- MHC-II- cells. Statistical analysis was performed using student’s t-test. (G) Absolute numbers of cardiac macrophages in hearts of uninfected-, passively immunized (αPly)- and naïve- mice infected with HIP or BIP 30 hours post infection. Macrophages were identified as CD64+MerTK+F4/80+CD11b+ cells. Statistical analysis was performed using student’s t-test. P value: * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001; data are represented as mean ± SEM.