The Streptococcus pneumoniae pilus-1 displays a biphasic expression pattern

Abstract

The Streptococcus pneumoniae pilus-1 is encoded by pilus islet 1 (PI-1), which has three clonal variants (clade I, II and III) and is present in about 30% of clinical pneumococcal isolates. In vitro and in vivo assays have demonstrated that pilus-1 is involved in attachment to epithelial cells and virulence, as well as protection in mouse models of infection. Several reports suggest that pilus-1 expression is tightly regulated and involves the interplay of numerous genetic regulators, including the PI-1 positive regulator RlrA. In this report we provide evidence that pilus expression, when analyzed at the single-cell level in PI-1 positive strains, is biphasic. In fact, the strains present two phenotypically different sub-populations of bacteria, one that expresses the pilus, while the other does not. The proportions of these two phenotypes are variable among the strains tested and are not influenced by genotype, serotype, growth conditions, colony morphology or by the presence of antibodies directed toward the pilus components. Two sub-populations, enriched in pilus expressing or not expressing bacteria were obtained by means of colony selection and immuno-detection methods for five strains. PI-1 sequencing in the two sub-populations revealed the absence of mutations, thus indicating that the biphasic expression observed is not due to a genetic modification within PI-1. Microarray expression profile and western blot analyses on whole bacterial lysates performed comparing the two enriched sub-populations, revealed that pilus expression is regulated at the transcriptional level (on/off regulation), and that there are no other genes, in addition to those encoded by PI-1, concurrently regulated across the strains tested. Finally, we provide evidence that the over-expression of the RrlA positive regulator is sufficient to induce pilus expression in pilus-1 negative bacteria. Overall, the data presented here suggest that the observed biphasic pilus expression phenotype could be an example of bistability in pneumococcus.

Figure 1. Pilus components display a biphasic expression pattern.

A) TIGR4 bacteria were labeled with anti-RrgA, RrgB clade I, RrgC, BgaA (beta-galactosidase), PhtA (pneumococcal histidine triad protein A) or PspC (pneumococcal surface protein C, also known as CbpA,) primary antibodies (1∶400 dilution), and with FITC anti-mouse IgG secondary antibodies (1∶100 dilution). Bacterial staining was analyzed by flow cytometry (FACS-Calibur). Sera of mice immunized with PBS were used as negative control. B,C) TIGR4 bacteria were processed for immunofluorescence, stained with mouse anti-RrgB antibodies (1∶2000 dilution) (red) and with S. pneumoniae anti-capsular antibodies (Omniserum 1∶2000 dilution) (green). Imaging was performed with a confocal microscope. Scale bar is 4 µm in panel B and 1 µm in panel C.

Figure 2. Pilus expression ratio is not correlated with serotype, genotype or clade type.

Bacteria containing either PI-1 clade I (A), clade II (B) or clade III (C), were labeled with clade specific anti-RrgB antibodies (1∶400 dilution) and FITC anti-mouse IgG secondary antibodies (1∶100 dilution). Pilus-1 expression was then analyzed by flow cytometry (FACS-Calibur).

Figure 3. Stable separation of enriched high (H) and low (L) pilus-1 expressing sub-populations.

A) TIGR4 pilus-1 expression was revealed on single colonies by colony immunoblot using anti-RrgB clade I antibodies (green, black and red circles correspond to colonies displaying low, medium or high RrgB specific signal intensities, respectively). Bacteria recovered from the growth of different colonies were stained with anti-RrgB clade I antibodies and analyzed by flow cytometry (B). The bacteria expressing (Pil+) and non- expressing (Pil-) the pilus-1 are indicated in the L (green) and H (red) enriched sub-populations, and in the wt (black). H and L sub-populations were stained for immunofluorescence (C and D). Bacteria were incubated with mouse anti-RrgB antibodies (1∶2000 dilution) (red) and with S.pneumoniae anti-capsular antibodies (Omniserum 1∶2000 dilution) (green). Imaging was performed with a confocal microscope. Scale bar is 5 µm.

Figure 4. PI-1 encoded proteins are not expressed in RrgB negative bacteria.

WB analysis performed on whole bacterial lysates of TIGR4 H (H), L (L) or TIGR4 L depleted of RrgB positive bacteria (D), using polyclonal mouse antisera against RrgA, RrgB, RrgC (see High molecular weight ladders), SrtC-1, SrtC-2 ,SrtC-3 (see bands indicated by arrows) and SrtA (used as loading control).

Figure 5. Pilus-1 expression is regulated at the transcriptional level.

A) High and low pilus expressing sub-populations of strains TIGR4 (Clade I), 19F Taiwan 14 (Clade I), OREP4 (Clade I), 6B Finland 14 (Clade II) and 35B SME 15 (Clade III) were labeled with clade specific anti-RrgB antibodies (1∶400 dilution) and FITC anti-mouse IgG secondary antibodies (1∶100 dilution). Pilus-1 expression was then analyzed by flow cytometry (FACS-Calibur). B) Schematic representation of PI-1. C) Log₂ ratio values indicating the PI-1 genes differential expression in High vs. Low pilus expressing sub-populations in the five above mentioned strains, as measured by spotted DNA microarray analysis. The data are measures of relative gene expression during in vitro growth in liquid cultures. The values reported for each gene are the mean of all the spots and their replicates within the array and of two independent experiments (bars represent standard deviations). D) Absolute gene expression levels of PI-1 genes measured for TIGR4 high and low pilus expressing sub-populations by microarray hybridization. Absolute expression levels reported for each gene are the mean of all the spots and their replicates within the array and of two independent experiments (bars represent the obtained standard deviations).

Figure 6. RlrA expression in pilus negative (Pil-) bacteria induces pilus polymerization.

TIGR4 low pilus expressing bacteria transformed with pMU1328 (panels A-C), pMU1328-Pc-rlrA (panels D-F), with pMU1328-Pc-rrgB (panels G-I) or with pMU1328-Pc-srtC-2 (panels J-L) were processed for confocal microscopy immuno-fluorescence analysis by incubating S. pneumoniae with anti-capsular antibodies (Omniserum 1∶2000 dilution) (green, left panels) and mouse anti-RrgB antibodies (1∶2000 dilution) (red, central panels). Right panels represent the merged signal of the left and central panels. Scale bar is 4 µm.