Temperature shift could serve as the cue for the commensal to invasive switch for S. pneumoniae

Cellular adhesion and invasion are pivotal to pathogenic infections and since the PCho-PAFR interaction is a key driver of invasive SPN infection, we reasoned that PCho levels in SPN might be elevated in pathologic respiratory conditions, mirroring the increased PAFR, and thereby facilitating pneumococcal infection. Indeed, SPN isolated from the lungs of mice, in a pneumonia model, had significantly higher surface PCho levels than those isolated from nasopharynx in an asymptomatic colonization model (Fig 1A). We have previously demonstrated that short-term exposure of mice to high temperatures promotes the dissemination of SPN from the upper to the lower airways in the nasopharyngeal colonization model [34]. Under these conditions, we were able to isolate SPN from both nasopharynx and lungs of individual animals. PCho surface abundance was higher on SPN recovered from nasopharynx of mice that had undergone short-term exposure to 40°C than in those housed under a constant temperate of 21°C, although this difference was not significant (P = 0.09) (Fig 1A). Furthermore, lung SPN showed greater PCho abundance than paired populations isolated from nasopharynx of the same animal (Fig 1B), confirming that descent through the airways is associated with increased PCho levels. To determine the possible factors driving increased PCho expression in lower airways, we next examined PCho levels in SPN in vitro, under a range of conditions designed to capture aspects of the stressed respiratory environment. Immunoblotting assays demonstrated that SPN PCho levels were upregulated by ~3 fold at temperatures associated with a febrile response (39°C), as compared to a nasal temperature (33°C) or as compared to levels measured in response to respiratory stress signals like ATP or the secreted cytokine TNFα (Fig 1C–1F). The increase in PCho that we observed at febrile temperature was confirmed to be surface-associated by flow cytometry analysis (Fig 1G and 1H). Interestingly, higher PCho levels at the febrile temperatures commonly associated with respiratory inflammation correlated with improved (~2-fold) cellular invasion ability of SPN (Fig 1I), irrespective of chain length alteration (S1A and S1B Fig). Together, these findings indicate that SPN utilize a robust host physiological symptom associated with respiratory pathologies, for modulation of surface properties. Notably, the link between temperature change and surface modifications has potential to contribute to infection progression and the commensal to pathogen switch.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Pneumococcal expression of phosphorycholine in in-vivo and in-vitro is triggered by temperature shift.

A. Flow-cytometry quantification of PCho on pneumococci recovered from nasopharynx and lungs of D39-infected mice. Mice were infected with 1x10⁵ colony forming units (CFU) for colonization of nasopharynx or 2x10⁶ CFU for colonization of lungs and a group of animals with SPN colonized in nasopharynx were exposed to 40ºC for 10 min before and 20 min after intranasal inoculation. SPN were recovered at 24 h post-infection by nasal wash or bronchoalveolar lavage. SPN were stained with anti-PCho antibody and median fluorescence intensity (MFI) recorded by flow cytometry. B. Paired nasopharynx and lung bacterial counts from individual mice exposed to 40ºC for 10 min before and 20 min following intranasal infection with 1x10⁵ CFU D39. SPN were collected at 24 h post-infection and processed as described in A. Lines connect nasopharynx and lung samples from individual animals. C, E. Immunoblot of total PCho in SPN following mock treatment (-) or treatment with either, ATP (10 µM) or TNFα (100 ng/ml) (C) or exposure to different temperatures (33°C and 39°C), for 3 h until late log phase (E). Enolase served as loading control. D, F. Quantification of PCho levels following ATP or TNFα treatment (D) or exposure to different temperatures (F). In both cases, normalizations were performed for individual samples against respective enolase levels. Fold changes were calculated using mock treatment as control for E and levels of PCho at 33ºC for F. G-H. Representative dot plot (G) and graph of flow cytometry analysis (H) of SPN cultured at 33°C and 39°C and immunostained with anti-PCho or isotype control antibody. 5x10⁴ pneumococcal cells were analyzed and gated based upon their respective unstained controls at each temperature. I. Invasion efficiency of SPN in A549 cells when pneumococcal cells following growth at 37°C were treated with either 10 µM ATP or 100 ng/ml TNFα or exposed to different temperatures (33°C and 39°C) for 3 h pre-infection. SPN infections were performed with MOI ~ 50 and fold change in invasion efficiency was calculated as ratio of intracellular bacterial CFU of the test to that of control. Statistical significance was assessed by Mann-Whitney test (A, B); One-way ANOVA (Dunnett’s test) (D, F, H) and two-tailed unpaired student’s t-test (I). ns, non-significant; *P < 0.05; **P < 0.01; ****P < 0.001. Data are mean ± SD of 3 independent biological replicates.

https://doi.org/10.1371/journal.ppat.1013545.g001

Thermosensing behavior of the CiaRH system dictates PCho levels in SPN

In SPN, PCho biogenesis and surface decoration is controlled by the lic operon. We profiled temperature-dependent expression of licA, a gene controlling the uptake and processing of choline residues from the environment [25], as well as htrA, a heat-responsive chaperone and protease that is reported to be induced during heat stress, as a positive control [21,35]. qRT-PCR analysis showed significant increases in both licA and htrA expression (~2-fold and 4-fold, respectively) upon exposure to febrile temperature (S1C Fig). The pneumococcal lic operon is positively regulated by the two-component system CiaRH. To understand the role of ciaRH in temperature-dependent increase in PCho levels, we created a ciaRH deletion strain. The ΔciaRH mutant failed to grow at higher temperature (39°C) due to the involvement of CiaRH in heat stress management (S2A Fig). Also, licA transcript levels were reduced in the ΔciaRH strain (S2B Fig) at 37°C. We also observed significant downregulation (~33%) of PCho surface abundance upon ciaRH deletion which could contribute to the almost complete loss of cellular invasion ability (~90-fold reduction) in this strain (S2C–S2E Fig). Importantly, the abolishment of invasive ability was not attributed to any alteration in bacterial chain length in ΔciaRH (S2F Fig). Collectively, this points to the central role of CiaR in regulation of PCho levels and the resultant invasive capability of SPN.

Multiple two-component systems, including CiaRH, are partially regulated by a positive feed-forward loop that increases their expression upon signal stimulation. Since PCho expression is raised at higher temperatures (via positive regulation of the lic operon by CiaR), we hypothesized that CiaR levels may increase by thermoregulation to kick start the positive feedback loop or meet the demand of increased PCho biogenesis in response to temperature related stress. We therefore examined the transcript and protein levels of ciaR at 33 and 39°C, to determine whether levels are positively regulated as a function of temperature. Intriguingly, we observed that although CiaR protein levels increased ~3-fold at higher temperatures in WT SPN (Fig 2A and 2B), ciaR transcript levels remained unchanged between 33 and 39°C (Fig 2C). Moreover, we found that CiaR levels were elevated only when pneumococcus was exposed to febrile temperature (39°C), but not at normal body temperature (37°C) (S1D and S1E Fig). This pattern of elevated protein levels with no apparent change in transcript abundance might be explained by the presence of RNA thermosensing elements at the 5’-untranslated region (5’-UTR) of ciaR. Indeed, the secondary structure of the 5’-UTR of ciaRH, modelled using an in silico RNA structure prediction tool, mFold [36,37], suggested occlusion of the ribosome binding site (RBS) [38], implying presence of a putative RNA thermosensor (RNA-T) (Fig 2D). Thermosensing elements function by intra-RNA base pairing that leads to mRNA secondary structures occluding the RBS. To determine whether the ciaRH mRNA sequence alone was sufficient to confer thermoregulation of gene expression, we expressed the full-length gene along with native promoter sequence in E. coli. We consistently observed significantly increased abundance (~35-fold) of CiaR at febrile temperature (39°C) compared to 33°C (Fig 2E and 2F). Next, we generated GFP-based reporter strains carrying the promoter regions of ciaRH (P_ciaRH) or a housekeeping gene, enolase (P_enolase), from SPN. In the engineered E. coli strains, the normalized amount of GFP (P_ciaR/P_enolase) was found to increase ~6 fold at higher temperature (Fig 2G and 2I), demonstrating that the 5’-UTR sequence of ciaRH alone is sufficient to confer thermoregulation of a different ORF in a heterologous system. To further eliminate involvement of host-dependent factors in the thermoregulation, an in vitro (cell-free) protein synthesis system was adopted, using the same constructs, where we observed ~4-fold increased GFP levels at 39ºC, compared to 33ºC (Fig 2H and 2I). Similar increments in the CiaR protein levels were observed when the in vitro transcription-translation reaction was decoupled to generate first the RNA, followed by translation using equal amount of transcripts (S1F Fig). The structural topology for the 5’-UTR ciaRH sequence was substantiated by circular dichroism (CD) spectroscopy [39]. CD melting studies performed with the chemically synthesized 5’-UTR sequence indicated formation of a hairpin structure with a T_m of 34 ± 0.14°C (Fig 2J). This implied opening up of the ciaRH RNA structure during the transition of temperature from 33 to 39ºC, exposing the RBS and thereby promoting enhanced ribosome binding to facilitate improved translation. We therefore measured the binding affinities of the purified 30S ribosomal subunit for its cognate Shine-Dalgarno sequence (5’-AGGAGG-3’) in the 5’-UTR of ciaRH mRNA using microscale thermophoresis (MST). We observed enhanced binding of the ribosomes to the 5’-UTR of ciaRH at 39°C compared to 33°C (Fig 2K). The mean equilibrium dissociation constant, K_D, of the 30S ribosomal subunit, for the 5’-UTR of ciaRH was found to be 195 nM at 33°C, which reduced to 2.36 nM at an elevated temperature of 39°C (Fig 2L). This showcases higher affinity of the 30S ribosomal subunit for the 5’-UTR of ciaRH at febrile temperature. Together, these results demonstrate the presence of an RNA-T like cis-regulatory element in the 5’-UTR of ciaRH, which is pivotal for thermal sensing.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Thermosensing ability of ciaR mRNA regulates PCho expression with temperature fluctuation.

A. Immunoblot demonstrating CiaR levels in SPN exposed to 33°C and 39°C. Enolase act as a loading control. Two different protein concentrations (10 µg and 20 µg) for each sample were immunoblotted to confirm the phenotype. B. Bar graph showing densitometric quantification of multiple CiaR immunoblots, like as shown in ‘A’. C. Bar graph showing transcript abundance of ciaR in SPN at 33°C and 39°C. 1 µg of total RNA isolated from SPN grown at different temperatures was reverse transcribed to synthesize cDNA and equal amount of cDNA from each test sample was used for analyzing Ct values using 2^–∆∆Ct method. ciaR transcript levels were normalized to 16S rRNA and expressed as fold change compared to 33°C. D. Cartoon diagram of predicted secondary structure of 5′-UTR of ciaRH transcript depicting presence of Transcription Start Site (TSS), Ribosome Binding Sequence (RBS) and start codon in silico predicted by mFold tool. E. Immunoblot showing ectopic expression of CiaR under its native promoter in E. coli at 33°C and 39°C. β-lactamase was used as a loading control to eliminate potential effects of temperature on plasmid replication. F. Graph depicting densitometric analysis of ‘E’. G. Immunoblot showcasing levels of GFP expressed under promoters of ciaRH (P_ciaR) and enolase (P_enolase, constitutive) in E. coli at different temperatures. β-lactamase was used as a loading control. Two different protein concentrations (10 µg and 20 µg) for each sample were immunoblotted to confirm the phenotype. H. Immunoblot showcasing levels of GFP expressed under P_ciaR or P_enolase promoters by in vitro transcription and translation at 33°C and 39°C. 1 µg of recombinant plasmids harboring P_ciaR-gfp or P_enolase-gfp were used as starting material and the reactions were incubated for 1 h at respective temperatures before precipitating the total protein and immunoblotting with anti-GFP Ab. I. Densitometric analysis of normalized GFP expressions in ‘G’ and ‘H’. J. Normalized CD melting curve for a 40-mer oligonucleotide containing the hairpin loop from the 5′-UTR of ciaRH. The CD data was fitted in the Boltzmann function in Origin to calculate the melting temperature (T_m). K. Binding curve demonstrating fraction of bound of 30S ribosomal unit (1 nM) to 5′-UTR of ciaRH RNA (2 µM) and negative control RNA (25 µM) (in vitro transcribed) serving as ligand at 33°C and 39°C. L. Plot depicting the K_D values of purified 30S subunit of bacterial ribosome binding to in vitro synthesized 5’-UTRciaRH RNA acquired from binding curve plot in ‘I’. Statistical significance was assessed by two-tailed unpaired student’s t-test (B, C, F and I). *P < 0.05; **P < 0.01; ***P < 0.005. Data are mean ± SEM of 3 independent biological replicates.

https://doi.org/10.1371/journal.ppat.1013545.g002

SPN is classified into multiple serotypes based on variations of capsular polysaccharides and in all serotypes the CiaRH system plays a critical role in their lifecycle [40]. In-silico analysis indicates that irrespective of serotypes, the 5’-UTR of ciaRH remains perfectly conserved across a set of 336 SPN genomes, covering 39 serotypes (S1 Dataset). Pneumococcus is an atypical member of the mitis group of Streptococci, otherwise composed of commensal bacteria. Evidence from phylogenetic studies suggests that Streptococcus mitis, the closest relative of SPN, evolved from a virulent ancestor resembling SPN and then lost virulence genes to adopt a more commensal lifestyle [41–43]. Despite such changes, S. mitis and several other streptococcus species, including species with opportunistic pathogenic potential, such as S. pseudopneumoniae [44] and S. toyakuensis [45] were found to harbor a conserved 5’-UTR ciaRH sequence (S3A Fig and S1 Dataset). The S. mitis ciaR 5′-UTR was predicted to form a RNA-T like secondary structure, as predicted by mFold (S3B Fig). When investigated experimentally, CiaR levels in S. mitis showed ~15-fold increase at febrile temperature compared to nasal temperatures (S3C and S3D Fig). These results illustrate presence of a functional temperature-regulated CiaRH TCS in S. mitis that could contribute to its occasional infectivity in immunocompromised patients [46]. Alternatively, the CiaRH regulon in this species might facilitate non-virulent resilience to temperature change. CiaRH belongs to a conserved subclass of OmpR regulators, which are widely present among microbes [47]. We used pBLAST to look for CiaR-like systems in Neisseria meningitidis (Nm), another pathogen occupying the same respiratory niche. We found >33% identity for CiaR with a response regulator MisR. Since MisRS is one of the crucial regulators in the virulent lifestyle of Nm [48,49], we investigated the role of thermoregulation of this locus. The secondary structure of the 5’-UTR of misR, predicted using mFold, shows an occluded Shine-Dalgarno (SD) region, suggesting presence of a potential thermosensor similar to that of ciaRH (S3E Fig). Indeed, abundance of GFP, driven by the 5’-UTR of misR, increased significantly (>2 fold) at 39ºC compared to 33ºC, providing evidence for the presence of a putative RNA-T (S3F and S3G Fig). Importantly, invasive N. meningitidis displays PCho attached to its pili, whereas in commensal Neisseria spp. it is found on lipopolysaccharides (LPS) [50,51]. Although MisRS is reported to be involved in virulence and our findings suggest it to be a thermosensor, its role in regulation of PCho expression in Neisseria spp. remains to be explored.

PCho surface abundance and pneumococcal invasion efficiency are tunable by altering the permissiveness of ciaRH mRNA to translation

To further understand the role of ciaRH in temperature-dependent incorporation of PCho into the cell wall, we engineered SPN with altered ciaR thermosensing ability. Site-directed mutagenesis was performed to modify the original 5’-UTR of ciaRH to generate translation-restrictive (5’-UTR_ClosedciaRH) and translation-permissive (5’-UTR_OpenciaRH) SPN strains (Fig 3A and 3B). These engineered SPN strains, bearing more accessible (open) or restricted (closed) forms of the ciaRH 5’-UTR, relative to WT, displayed consistent CiaR levels at 33 or 39ºC (Fig 3C and 3D), indicating loss of thermal sensing, unlike their WT counterpart. Additionally, the level of CiaR was significantly increased for 5’-UTR_OpenciaRH at 33°C compared to that of WT and was found to be similar to the elevated levels of CiaR in WT at 39°C, whilst CiaR levels for 5’-UTR_closeciaRH was found to be ~ 2–3 fold lower, even at 39°C, with respect to WT (Fig 3D). We also generated GFP-based reporter strains carrying 5’-UTR_OpenciaRH and 5’-UTR_ClosedciaRH sequences. In the engineered E. coli strains, both of the variants showed consistent GFP levels at 33°C vs 39°C. The expression of GFP was significantly higher in the 5’-UTR_OpenciaRH strain compared to the 5’-UTR_CloseciaRH strain in both temperatures (S4A Fig). Growth patterns, expression of major virulence factors (SpxB and Pneumolysin) and chain length of these engineered strains were unaffected, relative to WT (S4B–S4D Fig). We also analyzed the binding efficiency of purified ribosome to these engineered in vitro synthesized 5’-UTRs by MST analysis. The dose-response curve indicated enhanced binding of ribosomes to 5’-UTR_OpenciaRH template over that of 5’-UTR_ClosedciaRH RNA template, at both the analyzed temperatures. No significant temperature-dependent change in ribosome binding was observed for either of the RNA templates, confirming loss of thermoregulation (Fig 3E). We also observed that the expression of licA transcript (S4E Fig) and surface as well as total PCho levels in SPN (Figs 3F, 3G and S4F) were ~1.5-2 fold increased in translation-permissive and ~2–3 fold reduced in translation-restrictive SPN strains, respectively, when compared with WT. Consequently, higher number of intracellular bacteria were detected for the 5’-UTR_OpenciaRH strain, compared to the 5’-UTR_ClosedciaRH strain, when they were used to infect lung epithelial cells (Fig 3H and 3I), with the 5’-UTR_OpenciaRH strain showing ~3-fold higher invasiveness (Fig 3J). These data substantiate that superior adherence and invasion of SPN is fostered by increased PCho levels.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Genetic modification of ciaRH thermosensing element disrupts PCho levels and pneumococcal invasion efficiency.

A-B. Cartoon representation of predicted secondary structures of 5′-UTR_ClosedciaRH (A) and 5′-UTR_OpenciaRH (B) by mFold. List of mutations introduced for creation of each variant is mentioned in respective images. C. Immunoblot demonstrating CiaR levels in SPN strains harboring either 5′-UTR_ClosedciaRH or 5′-UTR_OpenciaRH variants following exposure to 33°C and 39°C. Enolase used as a loading control. D. Densitometric analysis of CiaR in ‘C’ normalized to individual Enolase levels followed by CiaR level for WT at 33°C. E. Binding curve demonstrating fraction of bound ribosomes to 5′-UTR_ClosedciaRH RNA and 5′-UTR_OpenciaRH RNA serving as ligands over the range of concentrations at 33°C and 39°C. F. Flow cytometry analysis of PCho levels in WT, 5′-UTR_ClosedciaRH and 5′-UTR_OpenciaRH containing SPN strains. Engineered SPN strains were incubated at 33°C and 39°C while WT SPN was grown and incubated at 37°C. n ≥ 15000 cells were analyzed in three (n = 3) independent expts. and data from a single expt. is represented as histogram. G. Graph showing fold change in mean fluorescence intensities of flow cytometry data shown in ‘F’. Fold change is calculated with respect to that of normalized PCho intensity of WT SPN cultured at 33°C. H-I. Confocal microscopy image depicting invasion of 5′-UTR_OpenciaRH (H) and 5′-UTR_ClosedciaRH (I) SPN strains (red) in A549 cells (stained with anti-microtubule Ab, cyan). Scale bar, 10 μm. J. Bar graph comparing fold change in invasion efficiencies of 5′-UTR ciaRH mutated SPN strains in A549 cells. A549 cells were infected with SPN with MOI ~ 50 following growth at 37°C and fold change in invasion efficiency was calculated as compared to invasion efficiency of 5′-UTR_ClosedciaRH. Statistical significance was assessed by two-way ANOVA followed by Šídák’s multiple comparisons test’s (D) or Tukey’s multiple comparisons test (G) and two-tailed unpaired student’s t-test (J). *P < 0.05; **P < 0.01; ***P < 0.005. Data are mean ± SEM of maximum 3 independent biological replicates.

https://doi.org/10.1371/journal.ppat.1013545.g003

PAFR expression upregulation is critical for improved pneumococcal invasion

Along with increased PCho levels, a concomitant rise in the host cellular expression of PAFR, the cognate ligand of PCho, is key for efficient pathogenic invasion. To determine the importance of PAFR in this process, we carried out an invasion assay by blocking PAFR with antibody as well as downregulating its expression using gene-specific siRNA (S5A Fig). In both cases, we observed a ~ 50% reduction in invasion efficiency of SPN compared to isotype control antibody or control siRNA-treated A549 cells (S5B and S5C Fig). This suggests a dominant role of PAFR in facilitating pneumococcal invasion in epithelial cells over other receptors. Similarly, blocking PCho moieties from SPN lowered the invasion by ~74%, underlining the importance of the PCho-PAFR axis in pneumococcal invasion (S5D Fig). Of note, and in contrast to PCho, cell surface as well as total PAFR expression in host cells remained unaffected by increase in temperature (S5E, S5G and S5H Fig). However, treatment of lung epithelial cells with TNF-α, a key mediator of inflammatory cascades, led to a significant increase in PAFR expression (S5F, S5G and S5I Fig). Expectedly, TNF-α induced-PAFR expression upregulation facilitated ~3 and ~4-fold increase in WT SPN invasion at 33 and 39ºC, respectively (Fig 4A–4C) [52]. We then checked the invasion ability of strains with altered ciaRH 5’-UTR in the experimental scenarios when PAFR levels were either upregulated or downregulated. Both 5’-UTR_ClosedciaRH and 5’-UTR_OpenciaRH strains exhibited equivalent invasion efficiencies with TNF-α treatment in comparison to mock at 33 or 39ºC (Fig 4D and 4E). Comparable invasion efficiencies were also observed with the engineered strains upon siPAFR treatment at 33 or 39ºC (Fig 4F and 4G). These results indicate that regardless of enhancement or reduction in PAFR levels, elevated invasion at febrile temperature is largely controlled by PCho regulation by CiaRH, suggestive of an influential role of the CiaRH thermosensing element on pneumococcal invasiveness at higher temperatures.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Codependence of PCho and PAFR for optimum pneumococcal invasion at higher temperature.

A-B. Confocal microscopy image depicting invasion of SPN (red) in A549 cells (stained with anti-microtubule Ab, cyan) following mock treatment (A) or treatment with 30 ng/ml TNFα for 12 h (B). Scale bar, 10 μm. C-E. Graph depicting fold change in invasion efficiency of WT SPN (C); or SPN strains harboring either 5′-UTR_ClosedciaRH (D); or 5′-UTR_OpenciaRH (E) variants in A549 cells. SPN strains were exposed to 33°C and 39°C and A549 cells were treated with either vehicle or TNFα (30 ng/ml) for 12 h before infection with respective SPN strains at MOI 50. Fold change in invasion efficiency is calculated as ratio of intracellular bacterial CFU of the test to that of mock treatment for respective strains at 33°C. F-G. Plots depicting a fold change in invasion of 5′-UTR_ClosedciaRH (F) and 5′-UTR_OpenciaRH (G) SPN mutants in A549 cells treated with siControl or siPAFR. Fold change is calculated by comparing invasion efficiency with that of siControl of individual strains at 33∘C. Statistical significance was assessed by two-way ANOVA (Bonferroni’s test) (C, D, E, F, and G). ns, non-significant; *P< 0.05; **P < 0.01; ***P < 0.005. Data are mean ± SD of maximum 3 independent biological replicates.

https://doi.org/10.1371/journal.ppat.1013545.g004

Engineered pneumococcal strains exhibit different outcomes in in vivo infection models

To further determine the extent to which thermal regulation via the 5’-UTR of ciaRH contributes to virulence and infection outcomes, mice were infected with WT, 5’-UTR_ClosedciaRH or 5’-UTR_OpenciaRH at a dose that induces invasive pneumonia when the WT strain is used. Strikingly, whilst the 5’-UTR_OpenciaRH strain displayed comparable virulence to WT in survival analysis, the 5’-UTR_ClosedciaRH strain failed to establish any symptomatic infection (Fig 5A). SPN numbers in lungs and blood were high in all mice that succumbed to infection, whilst those surviving until the experimental endpoint had cleared infection or retained only low CFU in organs (Fig 5B and 5C). To determine whether the pronounced phenotype of the 5’-UTR_ClosedciaRH strain resulted from a failure to establish colonization of the lung, mice were infected and then tissues were collected at 24 h post-infection. Whilst those mice infected with WT or the 5’-UTR_OpenciaRH strain displayed high lung bacterial burdens and evidence of developing bacteremia, those infected with the 5’-UTR_ClosedciaRH strain had already cleared infection by this time point (Fig 5D and 5E). Histopathological assessment of lungs harvested at 24 h post-infection showed extensive inflammatory infiltrate and gross pathology in WT and 5’-UTR_OpenciaRH infection, with lungs from 5’-UTR_ClosedciaRH infected mice retaining normal lung architecture (Fig 5F–5H). To rationalize the reduced potential of 5’-UTR_ClosedciaRH strain in causing infections, we assessed its colonization potential in the lower temperature environment of nasopharynx in an asymptomatic upper airway colonization model. Colonization density in nasopharynx with the 5’-UTR_ClosedciaRH strain was comparable to that of WT and 5’-UT_OpenciaRH at 7 days post-infection, but improved relative to other strains at 14 days post-infection (Fig 5I). Together, the findings from the disease and colonization models suggested that the 5’-UTR_ClosedciaRH strain retained an ability to establish a niche within the host upper airways, but that its infectivity was reduced in the higher temperature environment of lungs. It is unclear whether this reduced fitness in lungs is attributable to lower levels of PCho, decreased expression of virulence factors under CiaRH control or a combination of these factors. Spontaneous transition to invasive disease is rare in the nasopharyngeal colonization model used here, but microaspiration of low numbers of bacteria into the lower airways is observed. Despite similar nasopharyngeal bacterial burdens, we observed lower numbers of bacteria in trachea and lungs of 5’-UTR_ClosedciaRH infected mice, relative to those infected with the 5’-UTR_OpenciaRH strain (S7A Fig). Correspondingly, inflammatory cytokines in both nasopharynx and lungs were lower in 5’-UTR_ClosedciaRH infected mice (S7B and S7C Fig), suggesting that inflammation is driven predominantly by bacteria in the lung or that the reduced expression of CiaRH-regulated factors in the 5’-UTR_ClosedciaRH strain limited inflammation in the upper airways.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Disruption of the ciaRH RNAT influences virulence and infection outcomes in mouse models of pneumococcal pneumonia and nasopharyngeal colonization.

A. Mice were infected intranasally with 2x10⁶ CFU of D39, 5’UTR_OpenciaRH and 5’UTR_ClosedciaRH and monitored for signs of disease over 78 h. Mice were culled when they reached pre-determined disease severity endpoints. ****P < 0.0001 in Kaplan-Meier survival analysis vs D39 and 5’UTR_openciaRH. B-C. Bacterial burden in lungs (B) and blood (C) at time of death or, for mice that survived infection (below dashed line), at 78 h post-infection. D-E. Bacterial burden in lungs (D) and blood (E) at 24 h post-infection with 2x10⁶ CFU of D39, 5’UTR_OpenciaRH and 5’UTR_ClosedciaRH. F-H. Histological analysis of lung tissues at 24 h post-infection with WT-D39 (F), 5’UTR_OpenciaRH (G) and 5’UTR_ClosedciaRH (H). Lung tissue was stained with hematoxylin and eosin to evaluate pathology. Scale bar, 200 µm. I. Bacterial burden in nasopharynx at 7 and 14 days post-infection in mice infected intranasally with 1x10⁵ CFU of different SPN strains. Statistical significance was assessed using log-rank test (A), one-way ANOVA with Tukey’s multiple comparison test (D, E) and two-way ANOVA with Sidak’s multiple comparison test (I). ns = not significant, *P < 0.05, **P < 0.01, ****P < 0.0001.

https://doi.org/10.1371/journal.ppat.1013545.g005

As the PCho-PAFR axis contributed to the increased cellular invasiveness mediated by CiaRH (Fig 4), we sought to block this interaction as a means of limiting SPN disease progression. We therefore treated animals with phosphorylcholine intranasally, two hours prior to infection with a pneumonia-inducing dose of SPN. Pre-administration of PCho led to delayed kinetics of disease progression in WT D39-infected mice (median survival time 62 h), relative to untreated mice cohort (median survival time 40 h, p = 0.096) (S6A Fig). By contrast, PCho treatment did not influence infection outcomes for mice infected with the 5’-UTR_OpenciaRH strain. However, although no difference in virulence between WT D39 and the 5’-UTR_OpenciaRH strain had been observed in previous experiments (Fig 5), PCho treated mice infected with WT D39 (median survival time 62 h) showed a survival advantage relative to PCho treated mice infected with the 5’-UTR_OpenciaRH strain (median survival time 32 h) (p = 0.044) (S6A Fig). Bacterial burden in lung and blood samples from 24 h post-infection with both WT D39 and 5’-UTR_OpenciaRH strain suggested that the delayed disease progression in PCho treated animals may have resulted from delayed seeding into blood (S6B and S6C Fig). These effects likely result from a combination of PAFR blockade and the immunomodulatory effects of PCho on the host [53].

Ethics statement

Animal experiments were performed at the University of Dundee and University of Liverpool, with prior approval of the UK Home Office and the relevant Animal Welfare and Ethical Review Boards. Experiments were performed under project license PP2072053. The principles of the Declaration of Helsinki were observed throughout. Mice were housed in individually ventilated cages, with access to food and water ad libitum. Environmental enrichment was provided in all cages and mice were acclimatized to the animal unit for at least 10 days before use. Mice were randomly allocated to cages on arrival in the animal unit by staff with no role in study design. For experiments reported in this manuscript, individual mice were considered as the experimental unit. Sample sizes, controls and statistical analyses are detailed in the figures and accompanying legends.

Bacterial strains and growth conditions

Streptococcus pneumoniae (SPN) strain D39 (serotype 2, encapsulated) was kindly provided by Prof. E. Tuomanen (St. Jude Children’s Hospital, USA). The wild type (WT) and mutant strains were grown in C + Y synthetic media at 37⁰C in 5% CO₂. Glycerol stocks were prepared by mixing 900 µl of mid-exponentially grown SPN culture (OD_600nm ~ 0.4) with 600 µl of 80% sterile glycerol (32% final glycerol concentration) and stored in -80⁰C deep freezer. These glycerol stocks were used as starting inoculum for all experiments. The following antibiotics were used to grow SPN cultures when required: Spectinomycin (100 µg/ml) and Chloramphenicol (4.5 µg/ml). Escherichia coli (E. coli) DH5α was routinely grown in Luria Broth (LB) at 37⁰C with constant agitation (150–200 rpm). The following antibiotics were used for E. coli cultures when required: Ampicillin (100 µg/ml), Spectinomycin (100 µg/ml), Chloramphenicol (20 µg/mL), and Kanamycin (50 µg/ml). All bacterial strains used in the study are listed in S1 Table.

Cell culture

Human lung alveolar carcinoma (Type II pneumocyte) cell line A549 (ATCC No. CCL-185) were routinely cultured in DMEM (HiMedia, AT186) supplemented with 10% fetal bovine serum (FBS) (GIBCO, 38220090) at 37°C and 5% CO₂.

Mouse models of nasopharyngeal carriage and pneumonia

CD1 female mice (4–6 weeks old, Charles River, UK) were intranasally infected under light anaesthesia (oxygen and isoflurane) with 1x10⁵ colony forming units (CFU) in 10 μl PBS (nasopharyngeal colonisation model) or 2 × 10⁶ CFU in 50 μl PBS (pneumonia model). Disease severity was regularly monitored using a scoring system evaluating appearance, behaviour, breathing, and weight loss. Mice were euthanized via CO₂ and cervical dislocation either at pre-determined time points or when they reached severity endpoints. Blood was collected by cardiac puncture under terminal anaesthesia, and nasopharynx, trachea and lungs were harvested in PBS for CFU determination or 4% paraformaldehyde for histology. In some experiments, bacteria were isolated from nasopharynx by nasal wash and from lungs by post-mortem bronchoalveolar lavage using 100 μl and 1 ml PBS, respectively, containing 100 μg/ml gentamicin (Sigma). In these experiments, bacteria were subsequently stained with TEPC-15 (Sigma M1421), at a 1:100 dilution in 1% bovine serum albumin, to label PCho, for 1 h 30 min, under rotation. Bacteria were then washed once in PBS before addition of anti-mouse IgG, Alexa Fluor 488 (Invitrogen-A21202), at a 1 in 200 dilution in 1% BSA) for 30 minutes, under rotation. Nasopharynx sample staining was performed in 50 μl volumes, lung samples in 1 ml. For experiments aiming to block PAFR, mice were administered 2 mg phosphorylcholine oil (Cambridge Bioscience, UK), dropwise, under light anaesthesia, allowing each drop to be inhaled before adding the next. For determination of CFU in blood and organs, ten-fold serially diluted tissue samples were plated onto blood agar plates (Oxoid) containing 5% defibrinated horse blood and 40 μg/ml gentamicin (Sigma). Plates were incubated overnight at 37⁰C in an anaerobic jar, and bacterial colonies counted the next day.

Histology and imaging

Intact lung samples were rinsed in ice-cold PBS to remove residual blood and fixed in 4% paraformaldehyde, with transfer to 70% ethanol after 24 h. The tissues were then embedded in paraffin, sectioned with a microtome into 5 µm slices, and stained with haematoxylin and eosin. Tissue processing and imaging were performed by the University of Dundee Imaging Facility, with images captured using Axioscan 7 (Zeiss).

Cytokine quantification in tissue homogenates

Levels of mouse CXCL1 (KC) and TNF-α were determined in tissue homogenates from day three post-infection, in the nasopharyngeal carriage model. Undiluted nasopharynx samples and a 1:5 dilution of lung samples were used with the ELISA MAX Deluxe mouse CXCL1 and mouse TNF-α kits (Biolegend, UK), according to manufacturer’s instructions.

Identifying conserved 5′ untranslated regions proximal to ciaR orthologues across bacterial genomes

Bacterial species containing a ciaR gene were identified by running a nucleotide BLAST using the S. pneumoniae D39 ciaR nucleotide sequence against all bacterial genomes. For each hit, 300 bases upstream of the gene were extracted, and a local BLAST database was constructed from these sequences. The D39 5′ untranslated region (UTR) sequence, “AATAAGGAGGAGTTTCTC,” was then used as the query in a nucleotide BLAST against this upstream database to identify strains containing a similar UTR. The resulting matches were divided into two groups: S. pneumoniae genomes and other bacterial genomes. For the S. pneumoniae genomes, serotypes were assigned using PneumoKITy [68] (v1.0.1). For the other bacterial species, the most common UTR variant in each species was identified and selected as the representative sequence. This yielded one consensus UTR per species, which, together with the pneumococcal UTR, was aligned using Clustal Omega [69] (v2.3.0). The alignment was visualised and adjusted in Jalview [70] (v2.11.5.0). All custom bioinformatics scripts are available at: https://github.com/thomasebarton/Thermosensor-bioinformatics.

GFP based reporter strains in E. coli

All recombinant plasmids generated and primers used in the study are mentioned in S2 and S3 Tables, respectively. In order to assess the role of promoter, a gfp ORF (open reading frame) without start codon was amplified from another gfp containing plasmid and cloned within BamHI and XhoI restriction sites in pBluescript SK+ (pBSK+) to generate pAB309. Promoter regions comprising of 41 bp and 221 bp upstream of ciaRH operon and enolase (eno) gene, respectively, were amplified from genomic DNA of D39 were cloned in pAB309 to generate pAB414 and pAB372, respectively, to drive the transcription and translation of gfp. SacI and BamHI sites were used to clone P_ciaR, while XbaI and BamHI were used for P_enolase. Additionally, the 5’-UTR_OpenciaRH and 5’-UTR_ClosedciaRH was amplified from pAB377 (5’-UTRclose) and pAB376 (5’-UTRopen) (using primer set: pciaRopen-SacI-FP, pciaRopen-ATG-BamHI-RP, pciaRclosed-SacI-FP, and pciaRclosed-ATG-BamHI-RP) and were cloned in pAB309 using SacI and BamHI sites, to generate pAB1202 (5’-UTR_OpenciaRH -GFP) and pAB 1203 (5’-UTR_CloseciaRH-GFP).

Heterologous expression of CiaR in E. coli

For expression of CiaR in E. coli, ciaR gene along with its promoter (150 bp upstream of the start codon) was amplified from genomic DNA of D39 and cloned using XhoI and EcoRI restriction enzymes in pBSK+ to construct pAB382.

Construction of D39ΔciaRH

For deletion or insertion of modified genetic element/gene into native locus, allelic exchange by homologous recombination was employed. Regions (500 bp) flanking ciaRH operon and an chloramphenicol resistance cassette (Cm^r) amidst these regions were assembled in a pBSK+ vector (pIB168) [62] and used for pneumococcal transformation. A positive clone was confirmed by PCR using ciaRH-Flnk-F and ciaRH-Flnk-R primer sets and western blot (using CiaR antiserum, gift from R. Bruckner, University of Kaiserslautern, Germany).

Construction of D39 5′-UTR_Closed/OpenciaRH strain

For introducing mutations in 5’-UTR of ciaRH we used a previously made plasmid pAB368 [62] which possess both ciaR and ciaH ORF sequence flanked by upstream and downstream regions with a Spectinomycin resistance cassette (Sp^r) inserted in between. Site directed mutagenesis was employed to create nucleotide modifications in pAB368 (using primer set: closed-F, closed-R and open-F, open-R) to generate pAB377 (5’-UTRclose) and pAB376 (5’-UTRopen). Introduced mutations were confirmed by DNA sequencing and these plasmids were used for SPN transformation followed by selection of Sp^r colonies.

RNA isolation, cDNA synthesis and quantitative reverse transcriptase PCR

Total RNA from SPN culture was isolated using TRIzol reagent (ThermoFisher Scientific, 15596026) using RNA isolation kit (Qiagen,70104) as per manufacturer’s protocol. For RNA isolation, bacterial culture was homogenized in a bead beater in 1 ml TRIzol containing 0.1 mm glass beads. RNA was treated for 15 min with DNaseI (ThermoFisher Scientific,18047019) to remove any contaminating DNA and processed for reverse transcription. Approximately 5 μg of RNA was reverse transcribed using SuperScript First-Strand Synthesis System (ThermoFisher Scientific,18091050). The cDNA was further used for quantitative-PCR (q-PCR) for desired genes and 16S rRNA transcript abundance using iTaq Universal SYBR Green supermix (Biorad, 1725121). A relative change in the expression of gene was analyzed by 2^-ΔΔCt method.

Western blotting

Overnight grown E. coli cultures or mid-log phase grown SPN were re-suspended in lysis buffer (PBS containing 0.5 mM EDTA and 1 mM PMSF) and lysed by sonication. For lysing eukaryotic cells (A549s), monolayers of cells were washed several times with PBS and lysed in ice-cold RIPA buffer (50 mM Tris-Cl, pH 7.89, 150 mM NaCl, 1% Triton X-100, 0.5% Sodium deoyxycholate, 1% SDS, 10 mM NaF and 5 mM EDTA. The cell suspension was incubated at 4⁰C overnight and centrifuged to collect cell lysates. Crude extracts (both bacterial and eukaryotic cell lysates) were collected following centrifugation (15000 rpm, 30 min, 4˚C) and the total protein concentration was quantified using Bradford’s reagent (ThermoFisher Scientific, 23236). Approximately, 20 µg of total protein was separated on 12% SDS-PAGE gel and transferred to activated PVDF membrane (Biorad, 1620177) for 2 h in cold. Following transfer, the membrane was blocked in 5% skimmed milk made in TBST (Tris Buffer Saline, 0.1% Tween-20) for 2 h. Membrane was then probed with appropriate primary antibody solution followed by respective secondary antibody solution for 1 h. The immunoblot was developed using ECL solution (Biorad, 1705061).

Antibodies and reagents

Following antibodies were used in the study: Anti-Enolase serum, Anti-CiaR serum, were gifts from Prof. S. Hammerschmidt, University of Greifswald, Germany; J.N.Weiser, NYU Langone Health, USA; respectively. Antibodies specific for GFP (Invitrogen, A11122), His₆-tag (Invitrogen, MA1–21315), Pneumolysin (Santacruz, sc-80500), GAPDH (Millipore MAB374), TEPC-15 (Sigma M1421), anti-PAFR (Cayman chemical, 160602) were procured. The following secondary antibodies were used: HRP tagged anti-rabbit (Biolegend 406401), HRP tagged anti-mouse (Biolegend 405306), Biotin conjugated anti-mouse IgaA (Life Tech. M31115), Anti-mouse Alexa Fluor 488 (Invitrogen A21202), Anti-goat Alexa Fluor 633 (Invitrogen A21082). Recombinant Human TNF-α Protein (R&D systems 210-TA-020/CF) was used for induction of PAFR expression in A549 cells.

In vitro transcription/translation

To in vitro transcribe and translate reporter GFP protein, S30 cell free extract system (Promega, L1030) was used as per manufacturer’s protocol. Briefly, ~ 4 µg of plasmid DNA (pAB414 containing P_ciaR-gfp or pAB372 harboring P_enolase-gfp) was added in the mixture and incubated at 33⁰C and 39⁰C for 2 h. The reaction was then stopped by placing it in ice for 5 min. Additionally, to generate mRNA transcripts using in vitro transcription, DNA templates for P_ciaR-gfp and P_enolase-gfp were amplified using PCR from pAB414 and pAB372. RNA was produced from 1 μg of DNA utilizing T3 RNA polymerase (Promega, 2500 U, P4024) following the manufacturer’s guidelines. The RNA was then treated with Turbo DNase (Invitrogen, AM2238) to eliminate any leftover DNA template. Equal amounts of RNA (~10 μg) were subsequently translated using E. coli S30 Extract System (Promega, L1030), at different temperatures as per the manufacturer’s protocol. Total protein was precipitated by adding chilled acetone and centrifuged for 10 min at 4⁰C. Protein samples were prepared in 1X Laemmli buffer and equal amount was electrophoresed by SDS-PAGE. GFP expression was verified by immunoblotting with anti-GFP antibody.

Circular dichroism (CD) melting

CD melting studies were performed using the JASCO J-1500 CD spectrophotometer consisting of temperature control module. The spectra were obtained using 1 mm path length quartz cuvette at 260 nm from 5 to 95 °C temperature range, with a heating rate of 1 °C/min. 10 µM of CiaRH wild type (WT) RNA dissolved in 50 mM potassium phosphate buffer (33 mM KH₂PO₄, 17 mM K₂HPO₄), pH 6.5 was annealed by heating to 95°C for 5 min followed by gradual cooling to room temperature over 4–5 h. The sample was then kept at 4°C. The spectra were recorded and melting temperature (T_m) was evaluated by analyzing the CD spectra in Origin software (version 9.65) using the Boltzmann function.

Flow cytometry

For flow cytometry, mid-exponentially grown SPN cultures (OD_600nm ~ 0.4) were washed twice with PBS and 10⁸ bacteria were incubated with TEPC-15 antibody (anti-PCho) (1:100) for 60 min at room temperature (RT). Following two washes with PBS, bacteria were incubated with goat anti-mouse IgG Alexa Fluor 488 tagged secondary antibody (1:100) for 60 min at RT. Finally, bacterial cells were washed thrice with PBS and checked by flow cytometry using a BD-FACS Aria-Fusion flow cytometer utilizing forward and side scatter parameters. Results were analyzed using FlowJo software and data was represented as percent of gated bacterial population positive for fluorescence marker compared to control unstained cells.

Bacterial invasion assay

SPN strains grown till mid log phase (OD_600nm 0.4) in C + Y media at specified temperature were pelleted and re-suspended in 3 mL assay medium for infection of A549 monolayers. 500 µl of bacterial suspension was used for infection yielding a MOI in the range of 50–100. Following 1 h of infection, the monolayers were washed with PBS and incubated with assay medium containing penicillin (10 µg/ml) and gentamicin (400 µg/ml) for 2 h to kill extracellular SPN. Cells were then washed with PBS and lysed with 0.025% Triton X-100. Lysates were spread plated on Brain Heart Infusion (BHI) agar plates to enumerate viable SPN. Percent invasion was calculated as (CFU in the lysate/ CFU used for infection) X 100. Fold change in invasion efficiency was represented as the ratio of invasion of the test relative to that of control.

In certain set of experiments, cultured bacteria were treated with TEPC-15 antibody (anti-PCho) (1:100) for 2 h prior to infection or A549 cells were incubated with anti-PAFR antibody (Cayman chemical, 160602) (1:100) for 2 h or TNFα (30 ng/ml) for 8 h before infection with SPN.

siRNA directed gene knockdown

For RNAi mediated knock-down of PAFR (siPAFR, Dharmacon ON-TARGET SMARTpool L-005709-00-0005-20 pmol), A549 cells grown till ~70% confluency were transiently transfected with siPAFR or scramble (siControl) using Lipofectamine 3000 as per manufacturer’s instructions. 24 or 36 h post transfection, cells were processed for western blotting or bacterial invasion assays.

Pneumococcal chain length analysis

SPN D39 WT and mutant strains were grown in C + Y media until an OD_600nm of 0.4. Cells were washed with PBS and fixed by the addition of 4% PFA for 15 min. Following washing and resuspension in PBS, a smear was made on glass slides. The coverslip was mounted with the addition of VectaShield with DAPI. The samples were imaged using a Nikon Eclipse Ti epifluorescence microscope and images were analyzed using ImageJ software; 60- to 100-pixel area was given as the measurement for a single coccus.

Microscale thermophoresis (MST) to determine ribosome binding to 5’-UTR of ciaRH

For assessing the affinity of ribosomes to 5’-UTR of ciaRH, microscale thermophoresis was employed. First, RNA of 5’-UTR of ciaRH was synthesized using MEGAscript in vitro RNA synthesis kit (ThermoFisher Scientific, AM1334) as per manufacture’s protocol. Briefly, 1 µg of DNA template (bearing T7 promoter, 5’-UTR of ciaRH including the RBS till start codon ATG) was utilized to synthesize 22 bp RNA stretch using the kit. The reaction was mixed and incubated for 16 h at 37°C. Next, the reaction was treated 1 µl TURBO DNase and incubated for 15 min at 37°C. Finally, RNA was purified using sodium acetate precipitation and concentration was checked by Nano spectrophotometer.

For the interactions, purified ribosomal protein complex was mixed with 50 nM of Monolith NT His-tag labelling Kit Red Tris-NTA (L008; Nanotemper Technologies, Germany) for 30 min. Labeled 30S ribosomal unit was then mixed with increasing concentrations of 5’UTR of ciaRH RNA, in HEPES-based buffer (200 mM HEPES, 1 M KCl, 100 mM MgCl₂, 5% Glycerol and 1 mM ATP) and kept at 33°C or 39°C for 15 min. The sample was then loaded into standard treated capillaries and assessed by Monolith NT.115 (NanoTemper Technologies, Germany). The data was analyzed using MO Control software (NanoTemper Technologies Germany) to determine the affinity of the interaction.

Immunofluorescence

For immunofluorescence assay, A549 cells grown on glass cover slips were pre-treated with TNFα (as described above) and infected with different strains of SPN at MOI ~ 25 for 1 h followed by antibiotic treatment for 2 h. Post infection, cells were washed with DMEM, and fixed with ice-chilled Methanol at -20°C for 10 min. Further, the coverslips were blocked with 3% BSA in PBS for 2 h at RT. Cells were then treated with anti-microtubule (for host cell) and anti-enolase (for SPN) primary antibody in 1% BSA in PBS for overnight at 4ºC, washed with PBS and incubated with suitable secondary antibodies in 1% BSA in PBS for 1 h at RT. Finally, coverslips were washed with PBS and mounted on glass slides along with VectaShield without DAPI (Vector Laboratories) for visualization using a Laser Scanning Confocal microscope (LSM 780, Carl Zeiss) under 63X oil objectives. The images were acquired after optical sectioning and then processed using Zen lite software (Version 5.0.).

Statistical analysis

GraphPad Prism version 9 was used for statistical analysis. Statistical tests undertaken for individual experiments are mentioned in the respective figure legends. p < 0.05 was considered to be statistically significant. Data are presented as mean ± standard error (SEM) of at least 3 replicates unless otherwise stated.