Our animal procedures of this study were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Chang Gung University (Approval Number: CGU15-138), and the Committee recognizes that the proposed animal experiment follows the guideline as shown in the Guide for Laboratory Animal Facilities and Care as promulgated by the Council of Agriculture. Executive Yuan, ROC. BALB/cByJNarl mice were provided by National Laboratory Animal Center (NLAC), NARLabs, Taiwan. All animals were housed in an animal facility at 22°C, with a relative humidity of 55%, in a 12 h light/12 h dark cycle, with sterile tap water and food available ad libitum.

Pneumococcal isolates form Chang Gung Children’s Hospital (CGCH) were grown at 37°C in Todd–Hewitt broth supplemented with 0.5% yeast extract (THY) or on blood agar supplemented with 5% defibrinated sheep blood in the presence of 5% CO₂. S. pneumoniae is a biosafety level-2 microorganism. The experiments handling the bacteria should follow all appropriate guidelines and regulations. Escherichia coli was grown in Luria broth. Antibiotics were added at the following concentrations when requited: spectinomycin at 250 mg/L for S. pneumoniae and 100 mg/L for E. coli and chloramphenicol at 4 mg/L for S. pneumoniae and 30 mg/L for E. coli.

A549 (ATCC CCL-185) human lung epithelial carcinoma cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco) at 37°C in a humidified incubator at 5% CO₂ and subcultured twice a week. Adhesion assays using A549 were performed as described previously (Nguyen et al., 2014), with some modification. Briefly, on the day before the adhesion assay, A549 cells were seeded in 24-well plates (1 × 10⁵ cells/well) and maintained in culture medium at 37°C in a 5% CO₂ incubator. Mid-log phase (OD₆₀₀ = 0.4 to 0.6) S. pneumoniae in THY were washed once with serum-free DMEM medium and then added to each well at a multiplicity of infection (MOI) of 100 and co-incubated for 2 h at 37°C in a 5% CO₂ incubator. To determine the total numbers of adherent bacteria, the wells were washed three times with phosphate-buffered saline (PBS) to remove non-adhering bacteria, and 1 mL of deionized distilled water was added and incubated for 5 min to lyse cells. The adherent bacteria were enumerated by plating serial dilutions of the lysate on blood agar and counting the number of viable bacterial colony-forming units (CFUs). The experiments were performed in triplicate and repeated independently four times.

For Illumina sequencing, total RNA was extracted from each sample (serotype 19A ST320 and Taiwan 19F-14 ST236) using the RNeasy mini kit (QIAGEN) and then treated with RNase-free DNase I (QIAGEN) for 15 min according to the manufacturer’s protocols. The integrity of the isolated total RNA was checked using an Agilent Technologies 2100 Bioanalyzer. Then, cDNA libraries were prepared according to the manufacturer’s instructions (Illumina), and an ABI StepOnePlus Real-Time PCR System was used for quantification and qualification of the sample library. The genome and gene information for serotype Taiwan 19F-14 ST236 and serotype 19A ST320 were downloaded from the NCBI database (accession numbers: CP000921.1 and CP001993.1, respectively). The expression level of each gene, as determined by RNA-Seq, was normalized to the number of reads per kilobase of exon region per million mapped reads (RPKM). The cut-off value for determining gene transcription was determined based on the 95% confidence interval for the RPKM values of each gene.

An aliquot (400 ng) of total RNA from each bacterial strain was subjected to cDNA synthesis using SuperScript IV Reverse Transcriptase (Thermo Fisher). The cDNAs of HMPREF0837_10263, HMPREF0837_10266, dnaN, ychF and 23S rRNA were quantified using KAPA SYBR^® FAST qPCR Master Mix (KAPA Biosystems) and an ABI 7900 Real-Time PCR system. The cycling conditions for the quantitative real-time PCR (qPCR) were as follows: 50°C for 2 min and 95°C for 2 min, followed by 50 cycles of 95°C for 15 s and 60°C for 30 s, and after 50 cycles, melting curve analysis was performed from 60°C to 95°C, with a 2% ramp rate. Sequences of the primers used for RT–qPCR are listed in Table 1. The relative mRNA expression levels were calculated according to the ΔΔCt method, with normalization to 23S rRNA levels.

The population with excised prophage were determined by qPCR. The following oligonucleotide pairs were designed for the experiment: prophage-excised-specific oligonucleotides to amplify a 151-bp fragment (RT-A/RT-B) near the attB site on the chromosome; non-prophage-specific oligonucleotides designed to amplify a 72-bp fragment (RT-10621-F/RT-10621-R) of HMPREF0837_10621. First, 19A ST320 genomic DNA was used as a template to amplify target gene fragments by PCR using the primer pairs 10621-F/10621-R (724-bp, SP_10621, for the non-prophage genomic region) and A/B (2,290-bp, EC, for the excised chromosomal DNA, “attB”), which were then cloned into pJET1.2/blunt (Thermo Fisher) to create plasmid standards for each target (named 10621/pJET and EC/pJET, respectively). Each amplification was carried out in a 20 μl reaction containing KAPA SYBR^® FAST qPCR Master Mix on an ABI 7900 system. Serial 10-fold dilutions of the plasmid standard containing 0.1–1 × 10⁶ copies were used to establish a standard curve to evaluate the Ct V.S. log₁₀(copy number). To determine the changes in prophage excision over time, a culture of 19A ST320 was grown in THY at 37°C and 5% CO₂ and genomic DNA was extracted at different time points. Then, the percentage of cells with excised prophage at each time point were analyzed by qPCR and normalized to the standard curve. To estimate the cell population with excised prophage at different time points, the number of copies of excised-chromosomal target region was divided by the number of copies of SP_10621 at each time point.

The integrase (int), dnaN and ychF gene-deletion strain (19A-Δint, 19A-ΔdnaN and 19A-ΔychF) was generated by replacing the gene with a spectinomycin resistance gene cassette as follow and primer used are listed in Table 1. Briefly, coding regions and flanking fragments for int, dnaN and ychF form the serotype 19A ST320 were amplified by PCR using primer pairs: int-up-F/int-dn-R for int, dnaN-up-F/dnaN-dn-R for dnaN and ychF-up-F/ychF-dn-R for ychF, respectively. The resulting PCR products were cloned into a pGEM^®-T easy (Promega) plasmid. The coding regions of the int, dnaN and ychF were then removed by inverse PCR (iPCR) with primer pairs: int-inverse-F/int-inverse-R for Δint, dnaN-inverse-F/dnaN-inverse-R for ΔdnaN and ychF-inverse-F/ychF-inverse-R for ΔychF, respectively and ligated to a PCR amplified spectinomycin antibiotic cassette (spec-F/spec-R) form pDL278 plasmid to create a deletion construct vector. This vector was transformed into S. pneumoniae serotype 19A ST320 by Competence Stimulating Peptide-1 (CSP-1; GenScript), and transformants were selected on spectinomycin.

To construct a complemented 19A ST320 integrase deletion mutant strain (19A-Δint::int), a single copy of the integrase gene was inserted into the non-coding region between the HMPREF0837_12134 and HMPREF0837_12135 as follows. The integrase gene was PCR amplified from 19A ST320 by using the primer pair Pint-F/int-R, and then cloned into the pGEM^®-T easy plasmid to generate Pint::pGEM-T. A chloramphenicol (cat) antibiotic cassette from pKO3-Km plasmid was PCR amplified by using the primer pair Pcat-F/cat-R, and then sub-cloned into the SacII site of Pint::pGEM-T to generate Pint-cat::pGEM-T. Then, the primer pair Pint-F/cat-R was used to produce a Pint-cat PCR fragment. Next, a 2,501-bp fragment of the insertion site was PCR amplified by using the primer pair 12133-F/12136-R and cloned into pGEM^®-T easy to create p12133-12136::pGEM-T. A fragment generated by iPCR using the above plasmid and the primer pair 12134-5-inverse-F/12134-5-inverse-R was then ligated with the Pint-cat PCR fragment to form the complementation vector p12133-Pint-cat-12136::pGEM-T. The complementation strain (Δint::int) was created by transforming the p12133-Pint-cat-12136::pGEM-T plasmid into the 19A-Δint strain using CSP-1 and selecting on chloramphenicol.

An in vivo competition assay for S. pneumoniae infection in mice was conducted as described previously (Hsieh et al., 2013). Mid-log phase (OD₆₀₀ = 0.2–0.6) bacteria were combined at a 1:1 ratio (CFUs) in PBS, and the inoculating doses were confirmed by plating on blood agar plates. Then, 3-week-old female BALB/c mice were challenged intranasally (IN) with 50 μL of bacteria (containing 5 × 10⁶ CFUs/bacterial strain). After 7 days, the mice were sacrificed with CO₂, and nasal lavages fluid was collected. The CFUs recovered were determined by serial dilution and plating on blood agar. For the competition experiment between wild-type strain and the integrase deletion mutant, the recovered bacteria were distinguished by replica plating on blood agar containing 250 mg/L spectinomycin. One hundred randomly picked colonies were tested per mouse. A competitive index (CI) was calculated based on the ratio of the competing bacterial strains recovered by nasal lavage, normalized to the ratio of the respective bacteria in the inoculum = (output_{test strain}/output_{wild type})/(input_{test strain}/input_{wild type}).

Data are presented as means ± standard deviation (SD) from more than three independent experiments. Statistical significance was assessed by a two-tailed Student’s t-test. The mouse CI assay was analyzed by a one-sample Student’s t-test (one-tailed) using log-transformed CIs to determine if the indices were significant. A p value less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 15.0 for Windows (Statistical Package for Social Sciences, Chicago, IL, United States).

In a previous study, we demonstrated that the serotype 19A ST320 strain exhibited higher colonizing ability than its ancestral clone Taiwan 19F-14 (ST236) in mice (Hsieh et al., 2013). In this study, we demonstrated that the serotype 19A ST320 strain showed five-fold higher adherence to A549 cells than the serotype 19F ST236 strain (p < 0.001) (Figure 1).

The transcriptomic analysis revealed 801 genes that were differentially expressed between the serotype 19A ST320 and 19F ST236 strains, including 641 and 160 genes that were upregulated and downregulated in serotype 19A ST320, respectively (Supplementary Table S1 and Supplementary Figure S1). Of note, the mRNA levels of many genes located in the prophage region (∼18 kb) between HMPREF0837_10256 (dnaN, encoding DNA polymerase III beta subunit) and HMPREF0837_10294 (ychF, encoding a GTP-binding protein) were significantly higher in 19A ST320 than in 19F ST236 (Figure 2A). A RT-qPCR assay was then conducted on two genes within the prophage region (HMPREF0837_10263 and HMPREF0837_10266) to verify the RNA-seq results. RT-qPCR confirmed an increase in the mRNA levels of these two genes in 19A ST320 compared to the corresponding levels in 19F ST236 (data not shown).

There are 34 open reading frames (ORFs) located within the prophage region in 19A ST320 and 19F ST236 (Figure 2B). But, there was an additional transposase gene in the prophage region in 19F ST236. A 14-bp direct repeat sequence “5′-CCC TTT TTG TGT TA-3′” flanked the prophage region, which should be the insertion site of this prophage. The GC content of the approximately 18 kb prophage region is 38%. These two strains shared 99% nucleotide identity in this prophage region.

We hypothesized that there was increased spontaneous prophage induction in 19A ST320, which accounted for the difference in mRNA levels of genes within the prophage between 19A ST320 and 19F ST236. PCR with several combinations of primers was used to assess spontaneous integration and excision of the prophage region in 19A ST320. If the prophage could excise from the bacterial chromosome to form an epichromosomal circular structure, the PCR products would be observed using primer pairs A/B (2,290-bp) and C/D (901-bp) (Figure 2C). PCR with the primer pair A/D (1,914-bp) would yield a product when the prophage is integrated into the bacterial chromosome (Figure 2C). In our experiment, we found that the prophage excise from the genome and exist as a double-stranded circular DNA molecule in the serotype 19A ST320 strain by PCR; whereas the prophage was integrated in the genome of the serotype 19F ST236 strain (Figure 2C). A DNA sequence analysis of the PCR fragments produced by primer pairs A/B and C/D confirmed the ends of the circular phage DNA were connected via one of the 14-bp direct repeat sequence (5′-CCC TTT TTG TGT TA-3′) flanking the prophage region (C/D); the other one 14-bp direct repeat was left in the chromosome (A/B) when the prophage region excised from the serotype 19A ST320 genome. We also tested for excision activity in 10 other serotype 19A ST320 isolates and 8 serotype 19F ST236 isolates by PCR. The results showed that all serotype 19A ST320 isolates had spontaneous prophage induction. Two of the eight serotype 19F ST236 isolates did not harbor the prophage, whereas the other six isolates had the prophage, but no spontaneous induction was observed (Table 2). Another 19 clinical isolates belonging to different serotypes were also examined. The results showed that these 19 clinical isolates did not contain the prophage (Table 2).

Bacterial cells were harvested at different growth phases, from early log phase to stationary phase, we found that the percentage of the cell population with excised prophage increased with cell growth into late-log phase, but decreased at stationary phase. The average percentage of cells with excised prophage across all culture time points was 17.17 ± 3.19% (Figure 2D).

The prophage region of 19A ST320 harbors a putative integrase gene at the 3′ end (Figure 2B). The integrase gene was deleted to study whether it was responsible for prophage excision and integration. Following deletion of the integrase gene in 19A ST320, no prophage excision was detectable by PCR in 19A ST320 isogenic integrase mutant (19A-Δint) (Figure 3A). Complementation of the deletion with an integrase gene at another chromosomal location (19A-Δint::int) restored prophage excision (Figure 3A). Also, the mRNA expression level of genes (HMPREF0837_10263 and HMPREF0837_10266) within the prophage region in the integrase deletion mutant (19A-Δint) were significantly lower than that in 19A ST320 wild type (p < 0.05) (Figure 3B).

The growth curves and CFUs in different phases of the integrase deletion mutant and 19A ST320 wild type were similar (Figure 4A). We tested whether prophage excision is associated with adherence because there was a difference in adherence to respiratory epithelial cells between the serotype 19A ST320 and 19F ST236 strains (Figure 1). The isogenic integrase deletion mutant (19A-Δint) of serotype 19A ST320 showed decreased adherence to A549 cells compared to the wild-type strain (p < 0.002), whereas the integrase complementation strain (19A-Δint::int) showed restored cell adherence capacity that was similar to the wild-type strain (p = 0.8) (Figure 4B).

To reduce the variance caused by differences among mice, we performed a competition experiment in which mice were inoculated intra-nasal (IN) with equal CFUs of serotype 19A ST320 (19A-WT) and the isogenic integrase deletion mutant (19A-Δint) to assess the role of prophage excision on colonization. At 7 days post-infection, the isogenic integrase deletion mutant (19A-Δint) was sixfold less competitive than the parent serotype 19A ST320 strain in colonizing the mouse upper airway (CI = 0.16; p = 0.02; Figure 4C).

We created a 19A ST320 mutant devoid of the prophage region (19A-Δphage) to test whether the circular prophage DNA molecule directly modulates bacterial adherence. No difference was found in adhesion ability between the serotype 19A ST320 (19A-WT) and the prophage deletion mutant (19A-Δphage) (p = 0.9) (Figure 5A). This result indicated that the circular prophage, on its own, did not directly impact cell adherence. Next, we found that the mRNA expression level of ychF was different between strains 19A-WT and 19A-Δint (p < 0.05) (Figure 5B). Transcriptome analysis also showed that the mRNA level of ychF was higher in serotype 19A ST320 than in 19F ST236. To analyze the bacterial adherence ability, as compared to the wild-type strain, we generated a 19A ST320 ychF deletion mutant (19A-ΔychF). Deletion of ychF in the 19A ST320 strain did not impact growth (data not shown). However, 19A-ΔychF had a lower bacterial adherence ability than 19A-WT (p < 0.01) (Figure 5C). The -35-promoter sequence of ychF is rearranged when the prophage region integrates into or excises from chromosomal DNA of serotype 19A ST320 (Figure 5D).